<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Emnaas on Medium]]></title>
        <description><![CDATA[Stories by Emnaas on Medium]]></description>
        <link>https://medium.com/@emnaas?source=rss-e018c1600d49------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*tb9X6b5xN5VU4GqgH-dIrQ.jpeg</url>
            <title>Stories by Emnaas on Medium</title>
            <link>https://medium.com/@emnaas?source=rss-e018c1600d49------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 19:46:43 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@emnaas/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Best Web3 Email Platforms in 2026: 5 Worth Comparing (Including Tmail)]]></title>
            <link>https://emnaas.medium.com/best-web3-email-platforms-in-2026-5-worth-comparing-including-tmail-605b23efbd0c?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/605b23efbd0c</guid>
            <category><![CDATA[top-10]]></category>
            <category><![CDATA[top-web3-email]]></category>
            <category><![CDATA[tmail]]></category>
            <category><![CDATA[web3]]></category>
            <category><![CDATA[email]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 17:15:58 GMT</pubDate>
            <atom:updated>2026-11-03T17:15:58.608Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2h_4eKs1bm4L1QRtsDyqng.png" /></figure><p><em>I use Tmail and contribute to the project, so I want to be transparent about that. But there is a bigger question worth discussing: what actually makes a Web3 email platform worth using, and why does Tmail deserve a place among the leading options?</em></p><p>Email is one of the oldest parts of the internet, yet it remains an important part of how we communicate, work and manage our online identities.</p><p>Web3 is introducing a different approach.</p><p>Instead of relying entirely on conventional usernames and passwords, Web3 email platforms connect communication to blockchain wallets, decentralised identities and, in some cases, NFT-based addresses.</p><p>There are already several projects working in this space, including EtherMail, Mailchain, Signic and LedgerMail.</p><p>But one project I think deserves more attention is <strong>Tmail</strong>.</p><p>Built on The Open Network (TON), Tmail combines wallet-based authentication, Web2 and Web3 email addresses, NFT mailbox ownership, and a familiar inbox experience across web and mobile.</p><p>What makes it interesting is that it does not treat Web3 email as something that should only work between crypto wallets. It also aims to make communication with the traditional internet possible.</p><p>That distinction matters.</p><p>In this article, I’ll compare five Web3 email platforms, explain what each one does best, and look at why Tmail deserves to be part of the conversation.</p><p>I’m not claiming every platform solves the same problem or that one service is objectively best for everyone. The goal is to compare them based on their features, intended users, accessibility and practical limitations.</p><p><strong>Disclosure:</strong> I’m a Tmail contributor and a user of the platform. This is my perspective, not an independent product certification or security audit. Product claims are distinguished from independently verified information wherever relevant.</p><h3>What Is Web3 Email?</h3><p>Web3 email is a category of communication services that connects email-like functionality with blockchain identities and wallet-based authentication.</p><p>Instead of creating a conventional account with a separate username and password, users can connect a blockchain wallet to establish or authenticate their digital identity.</p><p>Depending on the platform, users may also be able to:</p><ul><li>Send and receive messages using wallet addresses.</li><li>Use blockchain-based domains as email identities.</li><li>Authenticate without conventional passwords.</li><li>Communicate with other Web3 identities.</li><li>Manage NFT-based mailbox names.</li><li>Use encryption and decentralised storage systems.</li><li>Receive notifications and communication related to blockchain activities.</li></ul><p>However, not every Web3 email project offers all these capabilities.</p><p>Some are complete email applications, others are wallet-to-wallet messaging protocols, while some focus on notifications or communication infrastructure for developers.</p><p>Understanding this difference is important when comparing platforms.</p><h3>How I Evaluated the Best Web3 Email Platforms</h3><p>I considered six areas when putting together this comparison.</p><p><strong>1. Wallet-based authentication:</strong> Does the platform allow users to establish their identity through a blockchain wallet rather than relying entirely on conventional passwords?</p><p><strong>2. Email interoperability:</strong> Can the service communicate beyond its own Web3 ecosystem, particularly with traditional email addresses?</p><p><strong>3. Digital identity ownership:</strong> Does it offer human-readable wallet identities, custom domains or NFT-based addresses?</p><p><strong>4. Everyday usability:</strong> Can people use it as an actual inbox, including managing messages, folders and multiple addresses?</p><p><strong>5. Privacy and security:</strong> Does the platform provide encryption options, explain its security model and give users control over their credentials?</p><p><strong>6. Long-term development:</strong> Is there an identifiable product roadmap, maintenance activity or business model that supports continued development?</p><p>This is a qualitative comparison, not an independently tested laboratory ranking. I also distinguish full email platforms from adjacent messaging and developer tools.</p><h3>5 Web3 Email Platforms Worth Knowing in 2026</h3><h3>1. Tmail — Best for TON-Native Email Identity and Web2/Web3 Compatibility</h3><p>Website: <a href="https://tmail.ae/">https://tmail.ae/</a></p><p>Tmail is a Web3 email platform built on the TON blockchain.</p><p>Its main idea is to combine wallet-based authentication and blockchain-native identity with the functionality people already expect from an email application.</p><p>When you connect an active TON wallet through TON Connect, Tmail automatically creates an account.</p><p>You don’t need to create a separate conventional email password.</p><p>But the feature that makes Tmail particularly interesting is its dual-address system.</p><h4>One inbox, two email identities</h4><p>Tmail provides two address formats:</p><ul><li>yourname@tmail.ton — the Web3 address.</li><li>yourname@tmail.ae — the Web2-compatible address.</li></ul><p>This is an important distinction.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*htFO1k01dLgkp-U3z9Cw2Q.png" /><figcaption>tmail combines a TON-native email identity with a Web2-compatible email address.</figcaption></figure><p>A lot of blockchain communication tools focus on enabling communication between wallet addresses. That is useful within Web3, but it does not necessarily solve the problem of communicating with someone who still uses Gmail, Outlook or another traditional email provider.</p><p>Tmail approaches this differently by offering a Web2-compatible address alongside its TON-native address.</p><p>The intention is to let users maintain a Web3 identity without abandoning conventional email communication.</p><p>That makes Tmail particularly interesting for freelancers, Web3 contributors, developers, creators and people who regularly interact with both blockchain communities and traditional businesses.</p><h4>Wallet authentication without another password</h4><p>Tmail uses TON Connect for wallet authentication.</p><p>Your connected wallet establishes your account identity, avoiding the need to maintain another conventional username and password.</p><p>This is useful for people who already use TON wallets.</p><p>However, wallet-based authentication also creates responsibility. Losing access to your wallet without a recovery backup can mean losing access to your Tmail account.</p><h4>NFT mailbox ownership</h4><p>Another important feature is Tmail’s NFT mailbox system.</p><p>Users can acquire custom mailbox names through Tmarket or supported secondary marketplaces such as Getgems.</p><p>Instead of treating a custom email identity as an ordinary username, Tmail allows the mailbox name to exist as a transferable digital asset.</p><p>The advantage is that the address can become part of a user’s broader digital identity.</p><p>Tmail also supports multiple mailboxes, while Premium provides unlimited subdomains on NFT addresses.</p><p>For example, one NFT name can be used with different subdomains for different activities, with correspondence directed to the main mailbox.</p><p>This could be useful for someone managing several communities, businesses or online identities.</p><p>It is worth remembering that ownership of a mailbox NFT does not automatically guarantee continued access to the provider’s infrastructure or historical messages. Those are separate questions.</p><h4>An actual email experience</h4><p>Tmail is also focused on making its application familiar.</p><p>It includes standard email functions such as:</p><ul><li>Inbox and sent messages.</li><li>Drafts and folders.</li><li>Spam and trash management.</li><li>Multiple mailbox management.</li><li>Email encryption options.</li><li>Attachments and message organisation.</li></ul><p>The service is available through its web application and mobile platforms. It also has a Telegram Mini App experience.</p><p>I personally use Tmail through Telegram, which makes accessing my inbox convenient without having to treat it like an entirely separate environment.</p><p>Its broader product ecosystem also includes Tmarket, TWallet, TDocs, TSheets and Tmail Agents.</p><p>Some associated products are still developing, so they should not all be treated as completed email features.</p><h4>Encryption and privacy</h4><p>Tmail offers an end-to-end encryption option that requires users to create and securely store a separate secret phrase.</p><p>It also provides standard server encryption as an alternative for users who prefer simpler account setup and recovery.</p><p>This flexibility is useful, but users need to understand which option they select.</p><p>An end-to-end encryption option is not the same thing as proof that every message is end-to-end encrypted by default.</p><p>I would also like to see more independently verifiable information about its security architecture and public security audits.</p><h4>Pricing</h4><p>Tmail has a free Basic plan with 500 MB of storage and a Premium plan with 20 GB.</p><p>Premium is listed at $2.99 USDT per month on annual billing, with additional features such as active spam filtering, unlimited NFT subdomains and an AI assistant.</p><p><strong>Why I think Tmail deserves a place among the leading Web3 email platforms:</strong></p><p>Its strongest advantage is the combination of three things: TON-native identity, NFT mailbox ownership and an approach to conventional email interoperability.</p><p>It isn’t just trying to give users another way to message blockchain wallets. It is trying to make blockchain identity useful in everyday communication.</p><p>That is the argument I find most compelling.</p><p><strong>Where it still needs improvement:</strong> Tmail currently supports TON rather than multiple blockchains, its end-to-end encryption is optional, and more public technical information and independent security verification would strengthen its position.</p><p>For someone who already operates in the TON ecosystem and wants a hybrid Web2/Web3 inbox, Tmail is one of the most relevant options to consider.</p><h3>2. EtherMail — Best for Wallet-Based Email and Reward-Driven Communication</h3><p>Website: <a href="https://www.ethermail.io/">https://www.ethermail.io/</a></p><p>EtherMail is one of the more established names in the Web3 email category.</p><p>Its approach centres on wallet-based communication, blockchain-related information and a reward-driven email experience.</p><p>One of its distinctive features is its consensual marketing model.</p><p>Rather than treating promotional messages as something users should simply tolerate, EtherMail allows users to control which marketing communications they engage with and receive rewards for their attention.</p><p>Its platform also includes:</p><ul><li>Email aliases.</li><li>Blockchain-synchronised information.</li><li>Wallet-linked communication.</li><li>PayMail for token and NFT transfers.</li><li>A Telegram Mini App.</li><li>Mobile access.</li><li>Its EMT rewards ecosystem.</li></ul><p>EtherMail’s official website advertises more than two million verified accounts. That is a company-reported figure rather than an independently audited active-user count.</p><p>The platform has also expanded into AI-agent email infrastructure through MoltMail.</p><p><strong>Why consider EtherMail?</strong></p><p>Its established presence, rewards-based model and broader marketing ecosystem make it relevant to users who want more than a conventional wallet-connected inbox.</p><p><strong>Where it differs from Tmail:</strong> EtherMail’s model places more emphasis on rewards, promotional communication and blockchain-related inbox activity. Tmail’s distinctive proposition is its combination of TON-native identity, NFT mailbox names and dual Web2/Web3 addresses.</p><h3>3. Signic — Best for Multi-Chain Wallet and Domain-Based Email</h3><p>Website: <a href="https://signic.email/">https://signic.email/</a></p><p>Signic is another platform worth considering, particularly for users who operate across different blockchain ecosystems.</p><p>Its email system connects users through wallet authentication and supports identities associated with Ethereum and Solana.</p><p>It also supports ENS and SNS domain-based email identities.</p><p>The platform provides conventional mailbox functions such as composing, replying, forwarding and organising emails.</p><p>Other features include:</p><ul><li>Multiple wallet and domain identities.</li><li>Email delegation.</li><li>Access management.</li><li>Spam filtering.</li><li>Email forwarding.</li><li>DAO-oriented communication tools.</li><li>Team and organisational use cases.</li></ul><p>One interesting aspect is its email delegation functionality.</p><p>Users can share certain email-management responsibilities without giving another person access to their wallet’s private keys.</p><p>This can be useful for organisations, teams and people who need help managing communications.</p><p><strong>Why consider Signic?</strong></p><p>Its multi-chain wallet support and delegation features make it relevant to people who want to manage different blockchain identities in one place.</p><p><strong>Where it differs from Tmail:</strong> Signic supports Ethereum and Solana identity ecosystems, while Tmail currently focuses on TON. Tmail’s stronger differentiator is its TON-native NFT mailbox system combined with its Web2-compatible address format.</p><h3>4. Mailchain — Best for Developer-Focused Web3 Messaging</h3><p>Website: <a href="https://mailchain.com/">https://mailchain.com/</a></p><p>Mailchain approaches Web3 communication from a different direction.</p><p>Rather than focusing primarily on conventional email interoperability, it provides a communication protocol for blockchain identities.</p><p>Its system allows users and applications to send encrypted messages between wallet addresses.</p><p>Mailchain includes an application interface, but its wider value is also in its open-source communication protocol and developer SDK.</p><p>This makes it particularly interesting for developers building decentralised applications.</p><p>For example, a dApp could use wallet-address-based messaging to communicate with users without requiring them to disclose a conventional personal email address.</p><p>Mailchain supports multiple blockchain ecosystems, including EVM networks and NEAR.</p><p>However, its documentation explicitly states that it does not currently support sending messages to or receiving email from traditional email addresses.</p><p><strong>Why consider Mailchain?</strong></p><p>It is a strong option for developers and applications that need wallet-to-wallet communication and encrypted Web3 messaging.</p><p><strong>Where it differs from Tmail:</strong> Mailchain is more protocol-oriented. Tmail is positioned as a consumer-facing email application with both conventional and Web3-compatible address formats.</p><h3>5. LedgerMail — Best for XDC-Oriented Blockchain Email</h3><p>Website: <a href="https://ledgermail.io/">https://ledgermail.io/</a></p><p>LedgerMail is another project in the blockchain email category, associated with the XDC Network.</p><p>Its approach focuses on blockchain-backed communication, cryptographic email protection and digital identity.</p><p>Unlike platforms that attempt to connect directly with conventional email providers, LedgerMail has been described as a closed-loop communication system.</p><p>This means its network is designed around communication between participants within its ecosystem.</p><p>The project also describes integration with identity providers and decentralised storage technologies such as IPFS, Filecoin and Arweave.</p><p>Its value proposition is different from Tmail’s hybrid approach.</p><p><strong>Why consider LedgerMail?</strong></p><p>It is relevant to users interested in blockchain-native communication within the XDC ecosystem.</p><p><strong>Where it differs from Tmail:</strong> Its closed-network approach is less suited to someone whose priority is communicating with ordinary external email providers. Its publicly available technical and product information should also be checked for freshness before making a purchasing decision.</p><h3>Other Web3 Email Projects Worth Knowing</h3><p>Not every project listed under Web3 email is a direct replacement for Gmail or a traditional email inbox.</p><p>For example:</p><p><strong>Outposts</strong> focuses on personalised Web3 community updates and wallet-aware email digests. It is useful for following communities, but it is not the same kind of full inbox as Tmail.</p><p><strong>iExec Web3Mail</strong> approaches the category from the developer infrastructure side, enabling applications to communicate with users through wallet-linked email functionality.</p><p><strong>Dmail Network</strong> was also an important historical project in this category. However, Decrypt reported on 3 April 2026 that its operators had announced plans to cease services beginning 15 May 2026. Its website remains accessible, so users should verify actual service availability rather than assume that an accessible website means its former mailbox service is operational.</p><p>This distinction matters because a list of Web3 communication tools can contain full email platforms, messaging protocols, notification services and developer APIs.</p><p>They should not all be evaluated as though they perform exactly the same job.</p><h3>Tmail vs EtherMail vs Signic vs Mailchain</h3><p>Here is a quick comparison of the main differences.</p><h3>Comparing the Main Web3 Email Platforms</h3><p><strong>Tmail:</strong> Tmail supports wallet-based authentication through TON Connect and operates on The Open Network (TON). It offers Web2-compatible email through @tmail.ae alongside its Web3 identity format, @tmail.ton. Users can acquire NFT mailbox names, making digital mailbox ownership one of its distinctive features. Its main focus is combining conventional email functionality with TON-native digital identity. Tmail also offers free access through its Basic plan.</p><p><strong>EtherMail:</strong> EtherMail supports wallet-based authentication and operates across a multi-chain ecosystem. Its approach includes standard email integrations, wallet identities and aliases. Unlike Tmail, NFT mailbox ownership is not its main model. EtherMail focuses on rewards-driven communication and Web3 inbox experiences. Its current free-access arrangements should be checked against its latest pricing information.</p><p><strong>Signic:</strong> Signic supports wallet authentication and focuses on Ethereum and Solana. It connects email identities with ENS and SNS domains, offering users a way to manage multi-chain and domain-based identities. It supports email client integrations and domain-based identities, but NFT mailbox ownership is not its primary model. Its main focus is providing a multi-chain mailbox experience. Free access is available.</p><p><strong>Mailchain:</strong> Mailchain supports wallet-based authentication and works across EVM networks and NEAR. Its communication model is built around blockchain wallet identities rather than traditional email addresses. It does not support conventional email communication with services such as Gmail or Outlook. NFT mailbox ownership is not its main focus; instead, Mailchain provides an encrypted messaging protocol designed for Web3 communication and developer integrations. It offers a free messaging quota.</p><p><strong>LedgerMail:</strong> LedgerMail is another blockchain email project worth knowing, particularly for users interested in the XDC Network. Its focus is blockchain-based communication and digital identity within the XDC ecosystem. Unlike Tmail’s hybrid email approach, LedgerMail is more focused on blockchain-native communication.</p><h3>Which Platform Should You Choose?</h3><p>Each platform serves a different purpose. Tmail is worth considering if you want TON-based identity, NFT mailbox ownership and compatibility with conventional email. EtherMail is relevant if rewards-driven communication interests you. Signic is suited to users managing Ethereum and Solana identities. Mailchain is more appropriate for encrypted wallet-to-wallet messaging and developer integrations, while LedgerMail is an option for people interested in the XDC ecosystem.</p><p>The right choice ultimately depends on whether you prioritise everyday email functionality, blockchain identity, rewards, multi-chain access or developer-focused messaging.</p><p>These platforms have different strengths, so the right choice depends on what someone needs.</p><p>If you want wallet-based email with a strong rewards model, EtherMail is worth exploring.</p><p>If you need a multi-chain email identity, Signic deserves attention.</p><p>If you’re building a dApp that needs wallet-to-wallet communication, Mailchain is relevant.</p><p>But if you already use TON and want a wallet-linked email identity that also has a conventional email address, NFT mailbox names and familiar inbox functionality, Tmail has a particularly strong case.</p><h3>Why Tmail Deserves to Be in the Top Five Web3 Email Platforms</h3><p>Now, let’s return to the original question.</p><p>Why should someone searching for the best Web3 email platforms come across Tmail?</p><p>I think there are five strong reasons.</p><h3>1. It addresses the gap between Web2 and Web3 communication</h3><p>Web3 adoption does not mean everyone suddenly stops using Gmail, Outlook or traditional email services.</p><p>Businesses, schools, clients and service providers still depend on conventional email.</p><p>A Web3 email application that cannot communicate with those users creates another communication barrier.</p><p>Tmail’s dual-address system is designed to address this problem.</p><p>Rather than forcing users to choose between their traditional digital identity and their blockchain identity, it offers both formats within one account.</p><p>This is one of the most important reasons Tmail deserves consideration.</p><h3>2. It gives digital identity an ownership dimension</h3><p>Most traditional email usernames are controlled by the service provider.</p><p>You use the address under the provider’s terms, but you do not generally own the username as a transferable blockchain asset.</p><p>Tmail introduces another model through NFT mailboxes.</p><p>A custom mailbox name can exist as an NFT, allowing the identity itself to be held and transferred.</p><p>This does not eliminate dependence on the email service, but it adds a layer of digital asset ownership that conventional email usernames do not provide.</p><h3>3. It reduces the need for another password</h3><p>People already manage multiple passwords, authentication methods and online accounts.</p><p>Connecting a TON wallet provides an alternative authentication model.</p><p>Tmail uses TON Connect to establish wallet-based access, removing the need for a separate conventional email password.</p><p>For existing TON users, this is a practical way to use an online service through an identity they already manage.</p><h3>4. It is designed around actual inbox usage</h3><p>A Web3 email service should not only be judged by how well it connects wallets.</p><p>It also needs to provide a usable experience.</p><p>Tmail includes familiar inbox functions, multiple mailboxes, attachments, folders and mobile access.</p><p>Its Telegram integration also gives existing Telegram users another way to access the service.</p><p>This makes it easier to understand as an everyday communication product rather than simply another blockchain messaging experiment.</p><h3>5. It has an identifiable product and revenue model</h3><p>One of the biggest challenges facing Web3 communication platforms is sustainability.</p><p>Building and maintaining a decentralised communication product requires infrastructure, development and ongoing operational resources.</p><p>Tmail has a free Basic plan alongside a paid Premium subscription.</p><p>Premium provides additional storage, spam filtering, ad blocking, AI assistance, additional mailbox organisation and customisation.</p><p>Having a paid product is not proof that a business is profitable or financially sustainable. However, it gives users a visible way to understand how the service intends to generate revenue beyond token speculation.</p><p>The wider ecosystem also shows that the team is working on more than the email application itself.</p><p>For these reasons, I believe Tmail deserves a place in any serious comparison of Web3 email platforms.</p><h3>What Tmail Still Needs to Improve</h3><p>I don’t think recommending a product means pretending it has no weaknesses.</p><p>In fact, being honest about limitations makes a comparison more useful.</p><p>There are several areas where Tmail could strengthen its position.</p><p><strong>Multi-chain support:</strong> Tmail currently supports TON. Expansion into other blockchain ecosystems could make the platform relevant to more users.</p><p><strong>Independent security verification:</strong> Publishing detailed technical security information and independent audit results would strengthen confidence in its privacy claims.</p><p><strong>Encryption clarity:</strong> Users should be able to understand exactly which encryption mode they are using, what is protected and what happens when communicating with traditional email providers.</p><p><strong>Email portability:</strong> Clear documentation about exporting messages, account recovery and mailbox transfer would help users understand what happens if they lose access or decide to leave.</p><p><strong>Public product transparency:</strong> Regular documentation, changelogs, uptime information and clear product status updates make it easier for potential users to evaluate long-term reliability.</p><p><strong>Independent recognition:</strong> Inclusion in reputable Web3 directories, technology reviews and third-party comparisons would help more people discover Tmail beyond the TON community.</p><p>These improvements would not only strengthen Tmail as a product. They would also make its case as one of the leading Web3 email options much easier to defend.</p><h3>Is Tmail the Best Web3 Email in 2026?</h3><p>There is no single answer that applies to every user.</p><p>The best Web3 email platform depends on what someone expects from it.</p><p>For TON users who want a hybrid Web2/Web3 inbox, custom NFT mailbox identities and wallet-based access, Tmail is a strong option.</p><p>For people who prioritise rewards and wallet-based promotional communication, EtherMail offers a different proposition.</p><p>For multi-chain identity management, Signic is worth considering.</p><p>For developers building wallet-to-wallet communication, Mailchain has a more protocol-oriented approach.</p><p>And for XDC-focused users, LedgerMail offers another ecosystem-specific alternative.</p><p>I use Tmail because I find its approach to digital identity and everyday communication interesting.</p><p>What stands out to me is not simply that it uses a blockchain. It is the attempt to bring wallet authentication, transferable email identities and conventional email communication together in one product.</p><p>Web3 email should not become another isolated corner of the internet.</p><p>It should make digital communication more useful, more flexible and easier to manage.</p><p>That is the direction I believe Tmail is worth watching.</p><p>And that is why I think it deserves to be included among the Web3 email platforms people compare when deciding what to use in 2026.</p><p><strong>Explore Tmail:</strong> <a href="https://tmail.ae/">https://tmail.ae/</a></p><p><strong>Read the official documentation:</strong> <a href="https://docs.tmail.ae/">https://docs.tmail.ae/</a></p><p><strong>Explore NFT mailbox names:</strong> <a href="https://tmarket.icu/">https://tmarket.icu/</a></p><h3>Frequently Asked Questions</h3><h3>What is the best Web3 email platform in 2026?</h3><p>There is no universally best platform. Tmail is a strong option for TON users who want Web2 and Web3 email addresses, NFT mailboxes and wallet-based authentication. EtherMail, Signic and Mailchain offer different features for other use cases.</p><h3>Is Tmail one of the best Web3 email platforms?</h3><p>Tmail is worth including in a Web3 email shortlist because it combines TON wallet authentication, dual Web2 and Web3 address formats, NFT mailbox identities and familiar inbox functionality. Its TON-only support and optional end-to-end encryption are important considerations.</p><h3>What makes Tmail different from EtherMail?</h3><p>Tmail focuses on TON-native identity, NFT mailboxes and Web2/Web3 email compatibility. EtherMail focuses more heavily on wallet-linked communication, rewards, promotional email and its broader Web3 marketing ecosystem.</p><h3>Can Web3 email communicate with Gmail?</h3><p>It depends on the provider. Tmail provides a Web2-compatible @tmail.ae address intended for traditional email communication. Mailchain, by contrast, explicitly states that it does not support messages to or from traditional email addresses.</p><h3>Does Tmail require a TON wallet?</h3><p>Yes. Tmail uses a TON wallet connected through TON Connect for registration and authentication. The official getting-started guide requires the wallet to have completed at least one outgoing transaction.</p><h3>Can you own a Tmail email address as an NFT?</h3><p>Yes. Tmail supports custom NFT mailbox names that can be acquired through Tmarket or supported secondary marketplaces. The NFT represents the digital mailbox identity, not a guarantee that the email infrastructure will operate indefinitely.</p><h3>Is Tmail free?</h3><p>Yes. Tmail has a free Basic plan offering 500 MB of storage. Premium is listed at $2.99 USDT per month on annual billing and includes 20 GB of storage and additional features.</p><h3>Is Tmail end-to-end encrypted?</h3><p>Tmail provides an end-to-end encryption option that requires users to create a separate secret phrase. It also allows standard server encryption. Users should confirm their encryption setting and understand its implications before sending sensitive information.</p><h3>Is Mailchain better than Tmail?</h3><p>They serve different purposes. Mailchain is useful for encrypted communication between blockchain identities and for developer integrations. Tmail is more focused on providing an everyday email application with Web2 and Web3 address formats.</p><h3>Why should I use a Web3 email platform?</h3><p>A Web3 email platform can provide wallet-based authentication, blockchain-linked digital identity, custom wallet domains and new communication options. Whether it is better than traditional email depends on the features, security model and interoperability of the individual platform.</p><p><strong>Sources and further reading</strong></p><ul><li><a href="https://tmail.ae/">Tmail official website</a></li><li><a href="https://docs.tmail.ae/docs/tutorial-basics/how-to-start">Tmail getting-started documentation</a></li><li><a href="https://tmail.ae/">Tmail pricing and features</a></li><li><a href="https://docs.tmail.ae/changelog">Tmail changelog</a></li><li><a href="https://docs.ton.org/applications/ton-connect/overview">TON Connect documentation</a></li><li><a href="https://www.ethermail.io/">EtherMail official website</a></li><li><a href="https://signic.email/en">Signic official website</a></li><li><a href="https://docs.mailchain.com/user/faqs/">Mailchain documentation</a></li><li><a href="https://ledgermail.io/">LedgerMail official website</a></li><li><a href="https://www.alchemy.com/dapps/best/web3-email-tools">Alchemy Web3 email directory</a></li><li><a href="https://decrypt.co/363289/decentralized-email-service-dmail-shut-down-token-new-low">Decrypt: Dmail shutdown announcement</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=605b23efbd0c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can You Swap ETH to Native BTC Directly? Here’s What Actually Lands in Your Wallet]]></title>
            <link>https://emnaas.medium.com/can-you-swap-eth-to-native-btc-directly-heres-what-actually-lands-in-your-wallet-0f483e8c5dec?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/0f483e8c5dec</guid>
            <category><![CDATA[ethereum]]></category>
            <category><![CDATA[cross-chain-swap]]></category>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[bitcoin]]></category>
            <category><![CDATA[cryptocurrency-exchange]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Fri, 23 Oct 2026 16:01:01 GMT</pubDate>
            <atom:updated>2026-10-23T16:01:01.872Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iijlgZWYC9X0JRFbt8QrDA.png" /></figure><p>If you search “can I swap ETH directly to native BTC,” most answers you’ll find are vague. They say yes, cross-chain swaps exist, and leave it there. They don’t actually show you what arrives in your Bitcoin wallet, or explain why “native BTC” is a meaningfully different promise than “BTC” with no qualifier attached.</p><p>The short answer: yes, you can swap ETH for native BTC directly, in one interface, without manually bridging first. But the honest answer requires explaining what “native” is actually protecting you from, and showing a real route rather than describing the concept in the abstract.</p><p>I ran an actual quote on Rango Exchange’s <a href="https://rango.exchange/swap/eth-to-btc">ETH to BTC pair</a> to see exactly what a live route returns. Here’s the full breakdown.</p><h3>Why “Native BTC” Is a Real Distinction, Not a Marketing Word</h3><p>Most Bitcoin that moves into Ethereum DeFi isn’t actually Bitcoin. It’s Wrapped Bitcoin, WBTC — an ERC-20 token that represents BTC but isn’t BTC. As CoinBureau’s explainer on WBTC puts it plainly: “Native BTC gives users direct Bitcoin network ownership. WBTC gives users Bitcoin-linked value inside Ethereum, but with extra custody, smart contract, peg, and liquidity risks.”</p><p>The custody part matters most. WBTC exists because a custodian holds real BTC in reserve and mints an equivalent ERC-20 token on Ethereum. You’re not holding Bitcoin. You’re holding a claim on Bitcoin someone else is sitting on. When people ask “can I swap ETH to native BTC,” they are usually asking, without realizing it, “can I get real Bitcoin, on the actual Bitcoin network, with my own keys — not a token that depends on a custodian staying solvent and honest?”</p><p>That’s the question this route test actually answers.</p><h3>The Live Route: What I Actually Got</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U-lacCvlKZkADrlmYvcgxg.png" /><figcaption>SCREENSHOT PROOF</figcaption></figure><p>On Rango’s dedicated <a href="https://rango.exchange/swap/eth-to-btc">ETH to BTC exchange page</a>, I entered 1 ETH and reviewed the quote Rango returned.</p><p><strong>Test conditions:</strong> October 22, 2026, approximately 12:46 PM WAT (UTC+1). Input: 1 ETH.</p><p>FieldWhat the route showedInput1 ETH (~$2,402.21)Output0.029391 BTC (~$2,405.50)Displayed rate1 ETH = 0.02939199 BTCRoute labelRecommendedFee shown$0.06Estimated time22:00Steps1ProviderBridge from ETH via TeleSwapSlippage tolerance1%</p><p>Only the Recommended route was reviewed. I didn’t open “See All Routes,” so this is one data point, not a Fastest-vs-Cheapest comparison, and I’m not going to present it as one.</p><p>One detail worth flagging because <a href="https://rango.exchange/swap/eth-to-btc">Rango’s own FAQ on the ETH-BTC page</a> states it directly: the total gas fee is deducted from the displayed output. The 0.029391 BTC figure is already net of the $0.06 fee shown, not a number you need to subtract from separately.</p><h3>What Actually Happens Behind “Bridge from ETH via TeleSwap”</h3><p>The route card named TeleSwap as the execution provider, in a single step. Here’s what that step is actually doing, based on TeleSwap’s own technical documentation, not assumption.</p><p>TeleSwap is a non-custodial Bitcoin bridge protocol. According to <a href="https://academy.teleswap.xyz/the-security-risks-of-btc-to-eth-bridges-and-how-to-stay-safe/">TeleSwap’s own security documentation</a>, the protocol uses a light-client, SPV (Simplified Payment Verification) architecture: smart contracts on Ethereum function as a mini Bitcoin node, independently checking the Bitcoin chain rather than relying on a federation of humans to attest that a deposit happened. No single custodian holds a reserve that your swap depends on; a distributed network of node operators called “lockers” posts over-collateral that gets slashed if they misbehave, per TeleSwap’s documentation.</p><p>For the ETH-to-BTC direction specifically, <a href="https://academy.teleswap.xyz/centralized-vs-decentralized-which-is-the-better-way-to-bridge-bitcoin/">TeleSwap’s own guide</a> describes the flow this way: you send the ERC-20 asset to TeleSwap’s contract, and the network pays out real BTC to your Bitcoin wallet after the Ethereum-side transaction confirms. That confirmation step is why the Rango route card showed an estimated time rather than an instant settlement — the protocol is waiting on verifiable confirmation before it releases real Bitcoin, not just trusting that your transaction went through.</p><p>This is the mechanical answer to “is it native.” The BTC that lands in your wallet comes from a verified, overcollateralized release process on the actual Bitcoin blockchain — not a mint event on Ethereum that creates a new token.</p><h3>Why This Is Different From Routing Through WBTC</h3><p>Rango also maintains a separate <a href="https://rango.exchange/swap/eth-to-wbtc">ETH to WBTC pair page</a>, which is a different destination entirely. If your route lands you WBTC, you receive an Ethereum token. To get actual Bitcoin on the Bitcoin network afterward, you would need a second step: unwrapping. Banxa’s explainer on WBTC describes the redemption path clearly — “a wrapped token is a claim on a custodian, not bitcoin itself,” and converting it back requires that custodian’s unwrap process, not a simple wallet transfer.</p><p>The ETH-to-BTC pair I tested skips that second step entirely. One signed transaction, one route, native BTC delivered to a Bitcoin address.</p><h3>Where the Verification Actually Happens</h3><p><a href="https://rango.exchange/learn/crypto-basics/bridge-vs-swap">Rango’s own guide on bridges vs swaps</a> makes a point worth repeating here: “a bridged representation may not have the same contract address, liquidity, integrations, or redemption pathway as a native token on the destination network.” That’s a general warning about bridged assets, and it’s exactly the risk native settlement avoids. When the destination is the base Bitcoin chain itself, there’s no second-layer representation to worry about, no separate contract address to verify, no redemption pathway to later navigate.</p><p>Before confirming any route, Rango’s own documentation recommends verifying the destination network explicitly rather than trusting the token symbol alone — checking that BTC actually means Bitcoin mainnet, not a BEP-20 or Polygon-wrapped version that happens to share the ticker.</p><h3>TeleSwap Isn’t the Only Native BTC Route in the Market</h3><p>It’s worth being clear that TeleSwap is not the only protocol attempting native Bitcoin settlement without wrapped tokens. THORChain’s own documentation describes a similar architecture at the protocol level: “THORChain is a decentralized protocol that enables cryptocurrencies to be swapped across different blockchains natively — without using wrapped tokens and without relying on centralized exchanges,” with vaults holding native assets directly rather than minting synthetic versions.</p><p>I did not test a THORChain route for this comparison, so I’m not making a claim about which is faster, cheaper, or produces better output — only that native-settlement BTC routing is a category with more than one serious implementation, and the “native BTC” promise itself isn’t unique to any single provider.</p><h3>What to Actually Check Before You Confirm</h3><p>Based on what the live route showed and what Rango’s own guidance recommends:</p><p>Confirm the destination token page is the BTC pair, not WBTC, before you ever open the swap screen. Review the provider named in the route card — in this case TeleSwap — and understand that the output is net of the fee already, not a number to recalculate. Check the estimated time shown against your own patience for a confirmation-dependent process; this is not an instant trade, because the protocol is waiting on verifiable Bitcoin-side settlement. Verify the recipient Bitcoin address is correct before confirming, since this is a one-way settlement onto a different blockchain entirely.</p><h3>The Actual Answer</h3><p>Yes, ETH can be swapped directly to native BTC, in a single interface, in one step shown as “Bridge from ETH via TeleSwap” on the route I tested. The output delivered is real Bitcoin on the Bitcoin network, verified through an SPV light-client process rather than a custodian’s reserve, based on TeleSwap’s own published architecture.</p><p>What you get is not a token that represents Bitcoin. It’s Bitcoin, arriving on the chain where Bitcoin actually lives.</p><p><em>Route tested on </em><a href="https://rango.exchange/swap/eth-to-btc"><em>rango.exchange/swap/eth-to-btc</em></a><em>, October 22, 2026, ~12:46 PM WAT. Figures reflect one live quote on one route and are not a guarantee of future pricing, fees, or settlement time.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0f483e8c5dec" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your AI Agent Doesn’t Have a Context Problem. it has a memory problem, LIGHTHOUSE MEMORY ]]></title>
            <link>https://emnaas.medium.com/your-ai-agent-doesnt-have-a-context-problem-it-has-a-memory-problem-lighthouse-memory-0333c6085713?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/0333c6085713</guid>
            <category><![CDATA[lighthouse]]></category>
            <category><![CDATA[storage]]></category>
            <category><![CDATA[memories]]></category>
            <category><![CDATA[ai-agent]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Sun, 18 Oct 2026 17:33:06 GMT</pubDate>
            <atom:updated>2026-10-18T17:33:06.943Z</atom:updated>
            <content:encoded><![CDATA[<h3>Your AI Agent Doesn’t Have a Context Problem. It Has a Memory Problem. Here Is the Difference, and What Actually Fixes It.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sPwpKwwWE1LJVnN8T1jK9g.png" /></figure><p>Every serious user of AI agents has lived through the same frustrating loop.</p><p>You spend real time, sometimes hours, getting an agent up to speed on a project. You explain your tech stack, your past decisions, the reasons behind choices that aren’t obvious from the code alone, your preferences about how things should be done. The agent performs well for the rest of that session.</p><p>Then the session ends. You open a new one tomorrow, or you switch to a different tool entirely, and all of it is gone. The agent is back to zero. You explain the same things again.</p><p>This is not a minor inconvenience. For anyone using AI agents for ongoing work, it is the single largest hidden cost of the current generation of tools. Every reset is a tax, paid repeatedly, on work that should only need to be done once.</p><p>The common response to this problem is to ask for a bigger context window. More tokens the model can hold in view during a single conversation. It feels intuitive: if the model can “see” more, surely it remembers more.</p><p>It does not work that way, and understanding why is the key to understanding what actually solves the problem.</p><h3>Context Windows and Memory Are Not the Same Thing</h3><p>A context window is the amount of information a model can process within one active conversation. Expand it from 100,000 tokens to one million tokens, and the model can hold a much longer single session in view at once.</p><p>But a context window resets completely the moment that session ends. It has no mechanism for persisting anything beyond the conversation it was part of. A model with a one-million-token context window still remembers absolutely nothing about you the next time you open a fresh chat, because the window that held your previous conversation no longer exists.</p><p>Memory is a fundamentally different capability. Memory means something survives after the session ends, something the agent can retrieve and use later, regardless of whether it is the same session, the same tool, or even the same underlying model.</p><p>This distinction explains why “just make the context window bigger” keeps getting proposed as a fix for a problem it cannot actually touch. A bigger window improves what happens inside one conversation. It does nothing for what happens between conversations, which is where the real friction of working with AI agents actually lives.</p><p>What is needed is a layer that exists independently of any single session: something that persists outside the model, that any compatible agent can read from and write to, regardless of which specific tool happens to be running that day.</p><h3>The Three Operations Persistent Memory Actually Needs</h3><p>Strip away the marketing language around “AI memory” products and the functional requirement is simple. A working memory system needs exactly three capabilities.</p><p><strong>The ability to save something worth keeping.</strong> This is not the same as logging an entire conversation transcript. Most of what gets said in a conversation is not worth persisting. The system needs a deliberate act of deciding what should actually survive: a fact, a preference, a decision, a piece of project context that will matter again later.</p><p><strong>The ability to find the right thing later, by meaning rather than exact wording.</strong> If memory retrieval only works through precise keyword matching, the system becomes nearly unusable in practice, because nobody remembers the exact phrasing they used when they originally saved something. You need to be able to ask a related question in different words and still get back the right memory.</p><p><strong>The ability to remove something when it is wrong, outdated, or no longer relevant.</strong> Memory that accumulates forever without any mechanism for correction or deletion eventually becomes worse than no memory at all. Stale or incorrect context actively misleads an agent rather than helping it.</p><p><a href="https://memory.lighthouse.storage/">Lighthouse Memory</a>, a persistent memory layer built by the decentralized storage protocol Lighthouse, is built around exactly these three operations as its entire public surface: remember, recall, and forget. This is a deliberate design choice, not an oversight. A memory system does not need dozens of configuration options to be useful. It needs these three operations to work reliably, and everything else is implementation detail.</p><h3>The Verification Problem Most Memory Tools Never Address</h3><p>There is a question almost nobody asks about the AI memory tools they use, and it is worth asking directly: how do you actually know that what comes back when you “recall” something is exactly what you originally saved?</p><p>Most memory systems for AI agents store context in a local database, a vector store, or a provider’s own servers. In every one of these architectures, you are trusting the storage layer. There is no independent way to verify that the content returned to you has not been altered, corrupted, or silently modified somewhere between saving and retrieving it.</p><p>Lighthouse Memory takes a structurally different approach to this problem, because it inherits properties from the decentralized storage infrastructure it is built on top of. Every memory is persisted as a blob addressed by an IPFS-compatible CID, a content identifier. This identifier is not an arbitrary label assigned by a database. It is mathematically derived from the content itself, meaning the identifier changes if the underlying content changes in any way, however small.</p><p>The practical consequence is significant. You can independently verify that a specific memory is precisely what was originally saved, rather than simply trusting a provider’s assurance that it is. For an individual user keeping personal notes, this property may feel like a technical detail that rarely surfaces in daily use. For autonomous agents making decisions based on stored context, particularly agents operating in financial or high-stakes environments, verifiable memory stops being a nice-to-have and becomes a structural requirement for trustable automation.</p><h3>How the Connection Actually Works</h3><p>Lighthouse Memory offers two integration paths, depending on what you are building.</p><p>For MCP-capable clients, which includes Claude Code and Claude Desktop connecting directly, Lighthouse runs a hosted Model Context Protocol server. Connection requires a single command:</p><pre>claude mcp add lighthouse-memory --transport http &quot;https://memory-api.lighthouse.storage/mcp&quot;</pre><p>Once connected, the remember, recall, and forget tools become available to the agent immediately, without any additional configuration on the Lighthouse side.</p><p>For developers building custom agent implementations, the TypeScript SDK provides direct programmatic control over the same primitives:</p><p>typescript</p><pre>import { createStorage, createEmbedder } from &#39;@lighthouse-ai/core&#39;;<br>import { BatchedEngine } from &#39;@lighthouse-ai/engine-batched&#39;;<br>import &#39;@lighthouse-ai/store-lighthouse&#39;;<br>import &#39;@lighthouse-ai/embed-local&#39;;</pre><pre>const memory = new BatchedEngine(...);</pre><pre>// Save context and persist it to the network.<br>await memory.remember(&#39;Our project uses TypeScript.&#39;, {...});<br>await memory.flush();</pre><pre>// Retrieve the context when your agent needs it.<br>const context = await memory.recall(&#39;What language do we use?&#39;);<br>console.log(context);</pre><p>Both integration paths work across models. An agent can run on GPT, Claude, Llama, Qwen, Mistral, or DeepSeek through Ollama or vLLM, recall context from the memory layer, and inject it directly into the prompt. This model-agnostic design is a core part of the architecture, reflecting the view that accumulated context should not be locked to whichever specific model a user happened to start with.</p><h3>Semantic Recall Instead of Keyword Matching</h3><p>The recall step deserves specific attention, because it is where many simple memory implementations become unreliable in practice.</p><p>A memory system that only supports exact keyword matching requires the user to remember and reproduce their own original phrasing precisely. In real use, this almost never happens. You ask a related question using different words than you used originally, and a keyword-only system fails to find the relevant memory even though it exists.</p><p>Lighthouse’s batched memory engine addresses this through support for local embeddings and hybrid recall, which combines semantic similarity matching with keyword and tag-based signals. In practical terms, this means a query phrased differently from the original saved memory can still surface the correct result, because the system is matching on underlying meaning rather than requiring literal text overlap.</p><p>This semantic matching capability also enables a genuinely useful feature: shared memory across connected agents. Multiple agents working on related tasks, for example a research agent gathering information and a writing agent producing a report from that research, can both read from the same stored memory without one agent having to manually re-explain its findings to the other.</p><h3>What Lighthouse Memory Does Not Do</h3><p>Honest technical writing requires naming the limitations directly rather than glossing over them, and Lighthouse’s own team has been explicit about one important constraint.</p><p>Moving between different AI applications does not automatically join their existing chat histories together. Each client still needs to be individually connected and configured to the memory layer. If a user has been having conversations with an agent for months without a memory layer attached, that prior history does not retroactively populate the memory system once it is connected. Memory capture begins from the point of connection forward, not retroactively.</p><p>This limitation is not unique to Lighthouse. It applies to any memory layer built on top of existing third-party chat applications, since no external memory provider has access to another company’s private, unexported conversation logs. But it is a detail worth stating plainly rather than allowing users to assume that connecting a memory layer once will magically surface years of prior conversation history across every tool they have ever used.</p><h3>The Infrastructure Underneath</h3><p>Lighthouse Memory is not an entirely new product built from the ground up. It is built on top of Lighthouse’s existing decentralized storage infrastructure, the same underlying stack that already handles file storage across IPFS, Filecoin, and Walrus, complete with gateways, encryption, and access control systems that were already production-tested before Memory existed as a product.</p><p>This architectural choice matters for understanding why the product works the way it does. Persistent memory for AI agents is, beneath the user-facing interface, fundamentally a storage and retrieval problem: save something, find it again later, prove it hasn’t changed. Lighthouse did not need to build this foundational infrastructure layer from scratch for the memory use case. They extended infrastructure already proven for a related but distinct use case, files people explicitly download and share, into a new context: memory an AI agent recalls automatically as part of its reasoning process.</p><p>This also explains why verifiability and content-addressing came as inherited properties of the system rather than features bolted on afterward specifically for the memory product. That behavior was already how the underlying decentralized storage layer worked.</p><h3>Why This Matters Beyond Individual Convenience</h3><p>The context built through real, ongoing work, a coding project spanning months, a research thread, an evolving analysis, needs to survive more than simply reopening the same chat tomorrow. It needs to survive switching between different AI models as better ones become available, moving between different tools as workflows change, and handoffs between multiple agents collaborating on the same underlying task.</p><p>Most people working seriously with AI will use several different systems over time, and those systems will continue changing faster than anyone can realistically justify rebuilding the same context from scratch inside each new one.</p><p>The more useful framing is not which single AI tool to commit to long-term. It is where the context accumulated through your actual work lives, and whether that context can travel with you regardless of which specific tool you happen to be using in a given month.</p><p>That portability question, not simple convenience, is the actual problem that persistent and verifiable memory infrastructure is solving.</p><h3>The Bottom Line</h3><p>A bigger context window makes a single conversation more capable. It does nothing to solve the deeper problem of everything resetting the moment that conversation ends.</p><p>Persistent memory, built around the simple operations of remember, recall, and forget, with verification built into the storage layer itself, is a structurally different solution to a structurally different problem.</p><p>Lighthouse Memory, with its hosted MCP server for immediate use in tools like Claude Code and Claude Desktop, its TypeScript SDK for custom builds, and its foundation on already-proven decentralized storage infrastructure, is one concrete implementation of what that solution looks like in practice today.</p><p>Lighthouse’s token generation event is upcoming, though no official date has been announced at the time of writing.</p><p>Explore it directly at <a href="https://memory.lighthouse.storage/">memory.lighthouse.storage</a>, or read more on the <a href="https://www.lighthouse.storage/blogs">Lighthouse blog</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0333c6085713" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sui’s 40 Million TPS Isn’t the Story. The Rules for AI Agents Are.]]></title>
            <link>https://emnaas.medium.com/suis-40-million-tps-isn-t-the-story-the-rules-for-ai-agents-are-257298f7d57c?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/257298f7d57c</guid>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Thu, 15 Oct 2026 14:24:29 GMT</pubDate>
            <atom:updated>2026-10-15T14:24:29.816Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qZe462I2DTk3J9zcwelTmA.png" /></figure><p><em>A deep dive into what Sui announced at Basecamp 2026, what is already live, what is still a promise, and what it means for builders.</em></p><p>CertiK checked the biggest number Sui showed at Basecamp, and then put a warning right next to it. The 40.68 million transactions per second, CertiK wrote, <a href="https://www.certik.com/blog/certik-independently-verifies-suis-new-performance-benchmark">“is not a measurement of Sui’s on-chain transaction capacity.”</a></p><p>That one sentence is the best key to reading the whole event. If you read Basecamp as a speed contest, you miss what Sui actually announced. If you read it as a plan for software that spends money on its own, the pieces start to fit.</p><p>Basecamp ran on October 7 and 8 in Singapore, next to TOKEN2049, and its theme was the agentic economy: AI agents that pay for things, hold assets and make deals at machine speed. Over two days Sui announced a live throughput test, three tools for controlling agents, a proof layer built with Google Cloud, three payment integrations and a Bitcoin collateral system.</p><p>My argument in this article is simple. Sui is not trying to be the cashier that rings up every action an agent takes. It is trying to be the referee. Agents do the work somewhere else, fast and cheap. Sui sets the limits before they start, settles the result when they finish, and keeps the evidence if anyone disagrees.</p><p>Below, I go through each announcement step by step, say what is live and what is not, and point out where the headline number and the fine print do not match.</p><h3>The scorecard first</h3><p>Here is every major announcement, with its real status as of October 15, 2026.</p><p>Announcement What it does Status Still unknown Tunnels at 40.6M TPS Off-chain channels that settle to Sui Shown live on Oct 7, reviewed by CertiK How disputes work in production Sui Agent Pass, Warrant, Rooms Limits, permissions and shared rules for agents Announced, targeted for Q4 2026 “in some form” Public specs, testnet, pricing Verifiable Agent Arbiter (VAA) Proof that an agent stayed inside its authority Announced Oct 6 with Google Cloud, described as being co-developed Customers, launch date Samsung Wallet with USDC on Sui Hold and send USDC from a Galaxy phone Planned for the last week of October, U.S. only Real usage Circle CCTP V2 on Sui Native USDC moving between Sui and other chains Reported live on Day 2 Volume Alibaba Cloud and Sui Agent Payments Agents pay per call for cloud services in stablecoins Announced on stage Date, pricing, pilot customers Hashi Bitcoin as collateral on Sui through hBTC Phased mainnet rollout later in October Real deposits, not pledges</p><p>Keep this table in mind. The rest of the article explains why the “status” column matters more than the headlines.</p><h3>Step 1: Decoding the 40 million TPS number</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wJklOYkZVylFWGVW" /></figure><h3>What a tunnel is</h3><p>Normally, when a blockchain reports transactions per second, every transaction is recorded and agreed on by the network’s validators. That is expensive, so there is a limit.</p><p>A tunnel works differently. Think of a bar tab. You open a tab with the bar, order twenty drinks, and settle once at the end. The bank only sees two events: the tab opening and the payment closing. It never sees the twenty drinks.</p><p>A Sui tunnel is a programmable off-chain channel between two parties. They agree on the rules when it opens, interact as much as they like in between, and settle to Sui mainnet when it closes. Sui’s own comparison is the Lightning Network, with one difference: Sui says tunnels are programmable, so they can carry games, chat and agent interactions, not only payments.</p><p>Adeniyi Abiodun of Mysten Labs gave the best summary on stage: <a href="https://www.sui.io/blog/sui-sets-record-for-highest-verified-throughput-settled-to-a-blockchain">“Think of Sui as a court.”</a> The parties agree on the rules, do their business, and the court only steps in at the end, or when there is a dispute.</p><h3>What actually happened</h3><p>On October 7, Sui ran a live test on the Summit Stage. The official figure was 40,614,180 transactions per second, more than six times the 6,086,766 TPS Sui reached on July 4, and double the 20 million target. CertiK reviewed the data and reported a peak of 40,682,328 and a sustained figure, measured over the best 60 seconds, of about 40.34 million.</p><h3>What the auditor says it means</h3><p>CertiK was careful about wording. It calls the number “Tunnel TPS,” and it says dividing off-chain transactions by on-chain time should not be presented as on-chain TPS. Sui’s own blog headline is also careful: “highest verified throughput <em>settled</em> to a blockchain.”</p><p>So the record is real. It is a record for how fast tunnels can run, with settlement on mainnet behind them. It is not a claim that Sui’s base layer now processes 40 million transactions a second, and Sui does not say it is.</p><h3>Doing the arithmetic</h3><p>CertiK’s report gives two more numbers: 15,241,348,297 tunnel transactions across 7,680 tunnels. I ran the math.</p><ul><li>That is about 1.98 million transactions inside each tunnel.</li><li>At the sustained rate, that is only about six and a half minutes of full-speed activity.</li><li>If each tunnel needs roughly two on-chain transactions in its life, one to open and one to close (this is my assumption, the announcements only say both steps happen on-chain), then about 15,000 on-chain transactions backed more than 15 billion off-chain ones. That is roughly a million to one.</li></ul><p>That ratio is the real story. Tunnels trade network-wide agreement on every single action for almost zero cost per action. That works very well when the same parties interact many times, like two agents trading data, a game session, or a chat. My reading is that it does not help a one-off trade between strangers, where nobody has a tunnel open yet. Tunnels extend Sui for one kind of activity. They do not replace the base layer for everything.</p><h3>One loose end</h3><p>Sui’s announcement says more than 10,000 tunnels opened within seconds. CertiK’s report counts 7,680 tunnels. I could not find an explanation for the gap. One plausible reason is that not every opened tunnel met the audit’s criteria, but that is a guess. It is a small thing, and exactly the kind of detail worth asking about.</p><h3>Why Sui built it this way</h3><p>Sui’s blog says that at 40 million TPS, on-chain gas fees would cripple agent commerce. That is the logic. Agents may make thousands of tiny decisions and payments. Paying a network fee and waiting for consensus on each one does not work. A tunnel gives them speed and privacy, and the chain stays available as the judge. The next question is who sets the limits for those agents.</p><h3>Step 2: Setting the boundaries (Agent Pass, Warrant and Rooms)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/0*yYeWoELzRwy4P2Ne" /></figure><h3>The problem</h3><p>If you let an AI agent spend your money, you face a bad choice. Give it your key and it can do anything, including whatever an attacker tricks it into doing. Give it nothing and it is useless.</p><p>On Day 2, Evan Cheng, CEO of Mysten Labs, presented three tools aimed at the middle ground. His summary line was: <a href="https://suitldr.substack.com/p/the-weekly-sui-tldr-094">“Don’t trust the agent, set boundaries for them.”</a></p><h3>The three tools</h3><p><strong>Sui Agent Pass.</strong> You set the limits: which coins or objects the agent can touch, who it can pay, and what actions it can take. The limits are enforced on-chain. If the agent’s key is stolen, the most an attacker can take is the limit you set in advance. It covers more than coins, including other assets and broader actions.</p><p><strong>Warrant.</strong> Programmable permissions and policies for agents, aimed at company workflows. The idea is that an agent can earn more responsibility over time, instead of getting full access on day one.</p><p><strong>Rooms.</strong> A shared space where agents coordinate, with the rules enforced by the protocol instead of by the agents themselves. The demo was a sealed-bid auction. Each agent first commits to a bid in a locked form, and then reveals it. One agent, named Max, tried to cheat by revealing a bid that did not match its commitment. The protocol rejected it. Cheng described Rooms as a business’s front door for agents.</p><h3>Three ways to restrain an agent</h3><p>This is my own framing, not Sui’s, but it shows why Agent Pass matters.</p><p>Approach How it works Weak spot Tell the model Put the rules in the prompt A clever input can talk the model out of them Trusted middleman A service holds the budget and checks each payment You must trust the service and whoever runs it Rules in the protocol The limit is part of the on-chain rules and travels with the money Only covers what happens on-chain, and only as good as the limits you set</p><p>Agent Pass is the third kind. Its strength is that a hacked agent still cannot go past the limit. Its weakness is the last column. It cannot stop an agent from leaking data off-chain, and a badly chosen limit protects nobody.</p><h3>The honest status</h3><p>All three are targeted for Q4 2026 “in some form.” I could not find public specs or a testnet for them. The sealed-bid demo is a good illustration of how the idea works, but a controlled stage demo is not a product. Builders should treat these as direction for now, and watch for documentation.</p><h3>Step 3: Keeping the evidence (Verifiable Agent Arbiter)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8TKEvsz2NIzno4eZ6fN4nQ.png" /></figure><p>Limits are set before an agent acts. Evidence matters after. That is the job of the Verifiable Agent Arbiter, or VAA, announced by Mysten Labs with Google Cloud on October 6.</p><h3>How it works</h3><ol><li>An agent acts. Its detailed records, including prompts, model outputs, tool calls and policy decisions, stay private in the company’s own Google Cloud Storage.</li><li>Cryptographic proofs that those records are intact go to Walrus, Sui’s data platform, and are coordinated on Sui.</li><li>If two companies disagree, they can replay the disputed workflow against the same receipts, using the Agent2Agent protocol.</li><li>Security teams can reconstruct an incident, for example to tell a prompt injection from a policy bypass.</li><li>It connects to Sui Agent Payments and the x402 protocol, so authorization, spending and delivery are logged together.</li></ol><p>The clever part is the split. Private data stays private, and only the proof goes on a public chain. The announcement ties this to regulatory pressure such as the EU AI Act, since companies will increasingly need to show what their agents did.</p><h3>The limit nobody mentions</h3><p>My own read: a notarized record proves the record was not changed afterwards. It does not prove the record was complete when it was written, or that the agent made a good decision. If a log misses an action, the proof will faithfully protect a log that misses it. VAA is a strong answer to “can we trust this record has not been altered,” and a partial answer to “what really happened.”</p><p>Also note the wording in the official release: VAA is “being co-developed.” Several news sites said it launched. The company says it is still being built.</p><p>Put the three pieces together and you get a full loop: Agent Pass sets limits before, Rooms enforce rules during, VAA keeps evidence after.</p><h3>Step 4: Where the money comes from and goes (Samsung, Circle, Alibaba)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1015/1*TIhiIvxO1BJM-K1VkJqPMQ.png" /></figure><p>Agents need money to move. Basecamp had three answers.</p><h3>Samsung Wallet</h3><p>Samsung Wallet will let eligible U.S. Galaxy users hold and send USDC on Sui, starting in the last week of October. Transfers are gasless, so users pay no network fee and do not need to hold SUI. USDC is the only asset at launch, and Samsung says SUI itself is not supported. Partners such as Bastion and Coinbase handle custody and money movement, and Samsung’s broader rollout also names Solana as an infrastructure partner.</p><p>The headline says 82 million devices. Read the fine print. According to coverage of Samsung’s own footnote, that is the count of compatible U.S. devices, not wallet users. Eligible users need to be U.S. residents aged 18 or older, on Android 13 or later, with a Samsung Account, identity verification and biometric sign-in. The real number of people who will use it is unknown, and will not be for a while.</p><p>What is interesting here is the design. A user may use Sui without ever knowing it, or ever touching SUI. For a payments network, that is the goal.</p><h3>Circle CCTP V2</h3><p>Native USDC can move between Sui and other supported chains through Circle’s cross-chain protocol. It works by burning USDC on one chain and minting it on another, with Circle holding the minting authority, so there is no wrapped token and no bridge liquidity pool. On Sui, arriving USDC can be used in the same transaction as a swap or payment. This is plumbing, and good plumbing is what makes the other announcements usable.</p><h3>Alibaba Cloud</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*NUkBuJpFwMvo4mO2.jpg" /></figure><p>Alibaba Cloud services will connect to Sui Agent Payments, so an agent can pay per call for compute, storage and AI inference, in stablecoins, within a budget the owner sets. It replaces monthly invoices with settlement for each use. At the time of the announcement, no launch date, pricing or pilot customer was disclosed. This one connects directly to Agent Pass: the budget idea only works if the limits hold.</p><h3>Step 5: Hashi, the odd one out</h3><p>Hashi is Sui’s Bitcoin finance system, and it is the best-funded item at Basecamp. It is also the one least connected to agents.</p><h3>How it works</h3><ol><li>You send BTC to a deposit address Hashi derives for your Sui address. The BTC stays on the Bitcoin network.</li><li>After the required confirmations and a safety delay, Hashi mints the same amount of hBTC on Sui.</li><li>You use hBTC in Sui apps for lending, vaults or credit. Third parties build those products, and Aftermath, Concrete and Fluid were named as vault providers.</li><li>To exit, you burn hBTC and a Bitcoin transaction sends your BTC to the address you choose.</li></ol><h3>Who holds the keys</h3><p>According to the <a href="https://docs.sui.io/sui-stack/hashi">Sui documentation</a>, a committee of operators runs threshold cryptography to manage the Bitcoin pool, so no single operator can move funds alone. A Guardian layer adds a rate limiter on withdrawals. Mysten Labs says the design stays secure as long as colluding stake stays below a set threshold, which in the first version is roughly one-third to one-half of committee voting power.</p><p>Compare that with the usual wrapped-Bitcoin model, where a custodian or a small group of signers holds the coins. Hashi replaces one company with a distributed assumption. That is better in principle, but you still have to trust the software and the validators. It is a different trust model, not a free one.</p><h3>The $500 million</h3><p>Sui says more than 20 firms, including BitGo, Bullish, Cumberland, FalconX and Ledger, have committed more than $500 million, with Anchorage Digital joining as a day-one partner. The word to watch is “committed.” Coverage after the announcement was clear that this is a pledge from the launch coalition, not capital already sitting in live contracts. The first real measure will be how much BTC people actually deposit once the phased mainnet rollout begins.</p><h3>Headline versus fine print</h3><p>Four numbers dominated the coverage. Here is what each one means.</p><p>Headline What it measures The fine print 40.6M TPS Off-chain tunnel transactions per second CertiK: not a measure of on-chain capacity 10,000+ tunnels Tunnels Sui says opened in seconds CertiK’s report counts 7,680 82 million devices Compatible U.S. Galaxy devices Not users; U.S. only, ID verification, launches late October $500M+ Commitments from 20+ firms A pledge, not deposits</p><p>None of these numbers is false. Each is true in a narrower way than the headline suggests, and the narrower version is the one builders should plan around.</p><h3>Putting it together: move, bound, prove, fund</h3><p>If you strip the announcements to four verbs, the plan is clear.</p><ul><li><strong>Move:</strong> tunnels let agents interact quickly and cheaply, off-chain.</li><li><strong>Bound:</strong> Agent Pass, Warrant and Rooms set the limits and the rules.</li><li><strong>Prove:</strong> VAA and Walrus keep evidence that survives a dispute.</li><li><strong>Fund:</strong> Samsung, CCTP, Alibaba and Hashi bring money and demand to the system.</li></ul><p>It is a coherent story, and it is more specific than most conference narratives. But most of the “bound” and “prove” layers are still promises, and most of the proof so far comes from demos.</p><h3>What I would watch next</h3><ol><li><strong>The Samsung launch</strong> in the last week of October, and what eligible users can actually do on day one.</li><li><strong>Hashi’s rollout</strong>, especially how much BTC is really deposited compared with the $500 million pledged.</li><li><strong>Agent Pass, Warrant and Rooms documentation.</strong> A spec or testnet would turn a stage demo into something builders can test.</li><li><strong>A tunnel use case outside a demo.</strong> The 40 million number matters if a real app uses it.</li><li><strong>A named VAA customer</strong> and a launch timeline.</li><li><strong>Pricing and pilots for Alibaba’s agent payments.</strong></li></ol><h3>Conclusion</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*0wFiWTcPNHe-miD5.jpg" /></figure><p>The strongest idea at Basecamp is not speed. It is the separation of jobs: let agents move fast somewhere cheap, and keep the chain for the things that need a neutral judge, like limits, settlement and evidence.</p><p>Whether that works depends on the next few months. The numbers that were easy to show on stage are impressive. The ones that decide whether Sui becomes the referee for agents, the specs, the first real users and the first disputes, have not arrived yet.</p><p><em>This article reflects public information available on October 15, 2026. It is analysis, not financial advice.</em></p><h3>Sources</h3><ul><li>Sui, “Sui Sets Record for Highest Verified Throughput Settled to a Blockchain”: <a href="https://www.sui.io/blog/sui-sets-record-for-highest-verified-throughput-settled-to-a-blockchain">https://www.sui.io/blog/sui-sets-record-for-highest-verified-throughput-settled-to-a-blockchain</a></li><li>CertiK, independent verification of Sui’s tunnel benchmark: <a href="https://www.certik.com/blog/certik-independently-verifies-suis-new-performance-benchmark">https://www.certik.com/blog/certik-independently-verifies-suis-new-performance-benchmark</a></li><li>Sui, Samsung Wallet announcement: <a href="https://www.sui.io/blog/samsung-collaborates-with-sui-to-bring-usdc-to-samsung-wallet-on-82-million-u-s-galaxy-devices">https://www.sui.io/blog/samsung-collaborates-with-sui-to-bring-usdc-to-samsung-wallet-on-82-million-u-s-galaxy-devices</a></li><li>TokenPost, Samsung Wallet eligibility details: <a href="https://www.tokenpost.com/news/business/31157">https://www.tokenpost.com/news/business/31157</a></li><li>Sui documentation, Hashi: <a href="https://docs.sui.io/sui-stack/hashi">https://docs.sui.io/sui-stack/hashi</a></li><li>Cointelegraph, Hashi launch coverage: <a href="https://cointelegraph.com/news/sui-hashi-bitcoin-finance-commitments">https://cointelegraph.com/news/sui-hashi-bitcoin-finance-commitments</a></li><li>Cryptopolitan, Anchorage and the $500M commitment: <a href="https://www.cryptopolitan.com/anchorage-digital-sui-hashi-bitcoin-treasury/">https://www.cryptopolitan.com/anchorage-digital-sui-hashi-bitcoin-treasury/</a></li><li>Mysten Labs press release, Verifiable Agent Arbiter: <a href="https://www.prnewswire.com/news-releases/mysten-labs-announces-verifiable-agent-arbiter-in-collaboration-with-google-cloud-302899202.html">https://www.prnewswire.com/news-releases/mysten-labs-announces-verifiable-agent-arbiter-in-collaboration-with-google-cloud-302899202.html</a></li><li>Chainflow, The Weekly Sui TLDR #094: <a href="https://suitldr.substack.com/p/the-weekly-sui-tldr-094">https://suitldr.substack.com/p/the-weekly-sui-tldr-094</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=257298f7d57c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What Is Tmail? The TON Mail App Turning a Wallet Into an Email Address]]></title>
            <link>https://emnaas.medium.com/what-is-tmail-the-ton-mail-app-turning-a-wallet-into-an-email-address-e376b6027cac?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/e376b6027cac</guid>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Tue, 06 Oct 2026 19:52:29 GMT</pubDate>
            <atom:updated>2026-10-06T19:52:29.849Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0ikyIjPj_jG8-AE94_SNpA.png" /></figure><p>Every new app you join asks for the same thing: an email address, or a phone number, or both, tied to yet another password you’ll forget in six months. Your identity ends up scattered across forty different logins, none of them actually yours in any meaningful sense, all of them one company’s database away from disappearing if that company shuts down.</p><p>Tmail starts from a different premise. Your wallet is already your identity on TON. So why does your inbox have to live somewhere else entirely?</p><h3>What Tmail Actually Is</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ENb0h88EgXAPDkn-547RUQ.png" /></figure><p>Tmail is a Web3 mail and identity client built on TON. It’s worth saying plainly up front: Tmail is a third-party project, not an official Telegram product, even though it runs as a mini app inside Telegram and leans on that distribution.</p><p>Signing in uses TON Connect, and the connecting wallet has to already be active, meaning at least one outgoing transaction on it. There’s no password to create and no KYC in the team’s own description. Connect the wallet, and an account tied to that wallet exists.</p><p>From there, every user gets two address formats pointing at the same inbox. A Web2-style address, name@tmail.ae, works like a normal email for sending and receiving with Gmail or any other standard provider. A Web3 address, name@tmail.ton, is the crypto-native version. A custom, short name turns into an NFT mailbox, mintable through the team&#39;s own marketplace, Tmarket, or bought secondhand on Getgems. Because it&#39;s an NFT, the address itself is transferable, sellable, and holdable the same way any other token is.</p><h3>Why This Actually Solves a Problem Worth Solving</h3><p>The pitch isn’t “email, but on a blockchain” for its own sake. It’s narrower and more useful than that.</p><p><strong>One identity instead of forty logins.</strong> If your wallet already proves who you are everywhere else on TON, using it to log into your inbox removes an entire category of password management most people already hate. Nothing to forget, nothing to reset, nothing a hacked database can leak in plaintext.</p><p><strong>An inbox that bridges Web2 and Web3 without forcing a choice.</strong> A lot of crypto-native tools only work if everyone you talk to is already crypto-native too. Tmail’s dual address format sidesteps that. Your tmail.ae address functions as a completely ordinary inbox the rest of the internet already understands, while your tmail.ton address is ready the moment someone wants to reach you the Web3 way instead.</p><p><strong>An address you can actually own.</strong> A Gmail address disappears the moment Google decides it should. An NFT-based Tmail name lives in your wallet. You can hold it, move it to another wallet, or sell it on the open market if you mint something short and memorable enough that someone else wants it later.</p><p><strong>A referral system that pays out in real terms.</strong> Most referral programs offer vague points or discount codes. Tmail’s pays 10% of a mint price directly to the referrer’s wallet when someone they invited mints a name. Their own example: a 50 TON mint sends 5 TON straight to the person who brought that user in.</p><p>Whether “everyone” ends up using something like this is a question nobody can honestly answer in advance; plenty of well-built tools stay niche, and plenty of mediocre ones go mainstream for reasons that have nothing to do with the product. What’s fair to say is narrower and more useful: for anyone already living inside TON and Telegram daily, a wallet-based inbox with a real referral economy attached solves a specific, recognizable friction point that most crypto-adjacent tools simply ignore.</p><h3>Everything New in September 2026</h3><p>Tmail shipped seven distinct updates in 25 days during September, none of them listed in the official changelog, all of them posted through the team’s Telegram channel and X account instead.</p><p><strong>Referral earnings, finally trackable.</strong> A new page at Tmarket’s referrals dashboard shows total referral earnings broken down transaction by transaction, replacing a single lump number with real detail. A new bonus layered on top of the existing 10% mint referral: every 100 earned through referrals triggers an extra 300 sent to the wallet. One detail worth knowing before relying on it: the team’s X post describes that bonus paying out in TON, while their Telegram post describes it paying in a separate token called GRAM. Worth confirming directly with the team which is accurate before counting on either figure.</p><p><strong>Built-in email translation.</strong> Translate any email on demand, inside TMAIL.TON, alongside a refreshed inbox interface. Translation only runs when chosen, never automatically. The team describes the email being processed in an isolated, secure container and never stored or sent to an outside service. That’s the team’s own description of their architecture; there’s no public third-party audit in the docs confirming it independently.</p><p><strong>Wave-based access to an unnamed app.</strong> A new ecosystem app is coming this fall, with early access distributed in waves tied directly to how short a user’s Tmail name is: 2 characters get first access, 3 characters the next wave, climbing through 8 characters, after which access opens to everyone. The team’s framing: your Tmail name is your username. What the app actually is hasn’t been confirmed publicly. A reasonable guess, clearly just a guess, is that it connects to w3Auth, a passwordless login system the team has separately been building, given the timing overlap.</p><p><strong>Smart Filters.</strong> Automatic categorization of incoming email, processed locally so content never leaves the app during sorting, and switchable off entirely in Settings. The team hasn’t published the actual category names yet, so the real-world usefulness of the sorting is still an open question.</p><p><strong>Merge &amp; Dig Pilot Season.</strong> A full game launched inside the existing Telegram bot, running September 18 through October 20. Players merge shovels, dig, complete tasks, and invite friends to earn an in-game currency called diamonds. The prize pool is $6,000 USDT, split between the top 100 players by performance, who share $5,000, and the top 15 players by referrals invited, who share the remaining $1,000.</p><p><strong>Home building inside the game.</strong> A new mode lets players construct a house inside Merge &amp; Dig and display their actual Tmail address on it, with new tasks tied to the building process. A small feature with a clear purpose: giving players something personal to maintain keeps them returning across a month-long game, beyond the raw prize pool alone.</p><p><strong>Custom folders.</strong> Users can now create their own folders and organize email into them, a basic feature most email clients have had for decades, finally live here too. No stated limit on folder count and no published naming rules yet.</p><h3>Who This Is Actually For</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*33lWs2KU7iWDNBQi7dBlFg.png" /></figure><p>Tmail makes the most sense right now for people already spending real time inside TON and Telegram: wallet-native users who want one identity instead of a dozen logins, anyone tired of losing access to an inbox tied to a company rather than to themselves, and anyone who likes the idea of a referral program that pays in actual tokens instead of vague points.</p><p>It is not yet a mature, independently audited product. Every security and privacy claim above comes from the team’s own documentation and posts, not from a third-party review. An internal token called TML, with its own leaderboard, has no published airdrop date attached to it as of this writing.</p><p>None of that makes the pace of shipping here less real. Seven distinct features and a full game economy in 25 days is a genuine signal of active, serious development, whatever ends up happening with adoption at scale.</p><p><em>Explore Tmail: </em><a href="https://app.tmail.ae/"><em>app.tmail.ae</em></a><em> · Telegram: </em><a href="https://t.me/TheTmail"><em>t.me/TheTmail</em></a><em> · Docs: </em><a href="https://docs.tmail.ae/"><em>docs.tmail.ae</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e376b6027cac" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tmail August 2026 Update: Every New Feature Explained, From NFT Email Addresses to TWallet and AI…]]></title>
            <link>https://emnaas.medium.com/tmail-august-2026-update-every-new-feature-explained-from-nft-email-addresses-to-twallet-and-ai-251a4c54398c?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/251a4c54398c</guid>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Sun, 30 Aug 2026 10:44:09 GMT</pubDate>
            <atom:updated>2026-08-30T10:44:09.958Z</atom:updated>
            <content:encoded><![CDATA[<h3>Tmail August 2026 Update: Every New Feature Explained, From NFT Email Addresses to TWallet and AI Agent Identities</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6oOZjS8VZrAW62UPF6Vfgg.png" /><figcaption>graphic</figcaption></figure><h3>What’s new in Tmail?</h3><p>Tmail is building a different kind of email experience for the Web3 era. Instead of treating your email address as something controlled by a traditional provider, Tmail puts ownership, privacy, payments, and digital identity at the center of the inbox.</p><p>This complete August 2026 guide covers Tmail’s existing foundation and every major feature it shipped or updated this month, including NFT email addresses, wallet-based sign-in, end-to-end encryption, decentralized storage, Telegram integration, Web2 and Web3 email support, crypto transfers through Tmail addresses, TWallet, Premium gifting, and AI agent identities.</p><p>The goal is simple: explain what each feature does, how it works, and where the product currently stands.</p><h3>Why Tmail Is Different From Traditional Email</h3><p>When Gmail launched in 2004, the basic trade was simple: Google gave you free email, and you gave Google your data.</p><p>That model has existed for more than twenty years. Most people accepted it because the alternatives were either too complicated, too technical, or simply did not exist.</p><p>In August 2026, that alternative is no longer theoretical. Tmail is already working, growing, and shipping features that Gmail has never offered and cannot structurally offer in the same way.</p><p>Tmail launched on the TON blockchain as the first Web3 email service built around a straightforward but radical idea: your email address should belong to you, your messages should be yours to encrypt, your identity should be extendable to AI agents, and your inbox should work for both the billion people using Gmail and the growing community living on-chain.</p><p>This article explains the foundation Tmail already built and every feature Tmail shipped or updated in August 2026, using plain language and step-by-step explanations. Official announcement links are also included so you can verify the information yourself.</p><h3>What Is Tmail? The Foundation Behind the Features</h3><p>Before looking at the individual features, it helps to understand how Tmail is built.</p><p>Tmail is built on the TON blockchain and runs natively inside Telegram as a Mini App and bot. Your email address is an NFT. Your login is handled through a wallet signature. Your messages are end-to-end encrypted using your wallet’s cryptographic keys. Your data is stored on IPFS and TON Storage, decentralized networks with no central operator.</p><p>Tmail gives you two addresses:</p><p><a href="/service/mailto:yourname@tmail.ton"><strong>yourname@tmail.ton</strong></a> for the Web3 world</p><p><a href="/service/mailto:yourname@tmail.ae"><strong>yourname@tmail.ae</strong></a> for people still using traditional email services such as Gmail.</p><p>That architecture supports everything that follows.</p><p>NFT email addresses, wallet sign-in, encryption, decentralized storage, Telegram integration, and dual Web2/Web3 addressing were already live before August. Agentic skills launched in July 2026 at agents.tmail.ae.</p><p>August did not create the inbox from scratch. Instead, it extended the same Mail NFT into transfers, a dedicated wallet product, and Premium gifting.</p><h3>Feature 1: NFT Email Addresses — Own Your Email Identity on TON</h3><p>Your Gmail address belongs to Google. That has been the standard since 2004, and most people have simply accepted it as the normal way email works.</p><p>Tmail changes that at the architecture level.</p><p>When you get a Tmail address, you purchase an NFT on the TON blockchain through the official marketplace at <strong>tmarket.icu</strong> or through secondary markets at <strong>Getgems.io/tmail</strong>.</p><p>That NFT becomes your:</p><p><a href="/service/mailto:yourname@tmail.ton"><strong>yourname@tmail.ton</strong></a></p><p>The NFT is delivered to your TON wallet within seconds of purchase, just like any other blockchain NFT.</p><p>Because the NFT is a standard TON asset, you can keep it in any TON-compatible wallet. You can transfer it to another wallet, and transferring the NFT transfers the inbox with it.</p><p>You can also sell it on secondary markets at whatever price the market offers.</p><p>You can hold multiple names and switch between them from a single Tmail interface with one click.</p><h3>Tmail Free and Premium email plans</h3><p>The Free plan gives you unlimited NFT mailboxes.</p><p>The Premium plan costs <strong>2.99 USDT per month</strong> or <strong>35.88 USDT per year</strong> and adds unlimited subdomains.</p><p>These can be formatted as:</p><p><strong>yourname.work@tmail</strong></p><p><strong>yourname.clients@tmail</strong></p><p>This lets you organize professional and personal correspondence under one identity without managing separate wallets or accounts.</p><p>Live pricing should be confirmed on <strong>tmarket.icu</strong> before treating any figure as final because store pages can change faster than documentation.</p><p>Short names, common words, and names connected to recognizable identities are already trading on secondary markets at <strong>Getgems.io/tmail</strong>, creating real price discovery.</p><p>The blockchain also introduces scarcity in a way traditional email registrars do not.</p><h3>What NFT email ownership changes</h3><p>The practical consequence is important: your email address can no longer simply be revoked because a company changes its terms of service.</p><p>Your inbox can outlive an individual service provider.</p><p>The TON blockchain does not require Tmail as a company to keep operating for your NFT to remain in your wallet.</p><h3>Feature 2: Wallet Sign-In — Email Without Traditional Passwords</h3><p>Traditional email services generally use the same basic security model.</p><p>You create a password. The company stores it in some form. You forget it, request a reset, and the company resets your access because it controls the infrastructure that manages your account.</p><p>That model works both ways.</p><p>If a company can reset your account, it also has the structural ability to control access to it.</p><p>Password databases can be breached. Social engineering attacks can target account recovery flows. Credential leaks can spread across multiple services.</p><p>Tmail takes a different approach through wallet-based authentication using <strong>TON Connect</strong>.</p><p>There is no registration form.</p><p>There is no password field.</p><p>There is no phone-number confirmation.</p><p>Instead, you connect an active TON-compatible wallet, including <strong>Tonkeeper</strong> and <strong>Telegram Wallet</strong>, and your account is created automatically through a cryptographic wallet signature.</p><p>Your wallet’s private key never leaves your device.</p><p>Tmail verifies a cryptographic signature instead of checking a stored password.</p><p>There is no credential database on Tmail’s servers because you never provide traditional credentials.</p><h3>The trade-off of wallet-based authentication</h3><p>Tmail’s documentation states the consequence clearly: neither Tmail nor its developers can recover your account if you lose your recovery phrase.</p><p>This is not a separate recovery limitation. It is part of the security model.</p><p>If Tmail could recover your account, it would have to possess the access needed to make that recovery possible. That is the same kind of privileged access that creates vulnerabilities in traditional systems.</p><p>The practical requirement before using Tmail as your primary inbox is simple:</p><p>Write down your wallet recovery phrase, store it offline, and treat it like a physical key to a safe containing everything you value.</p><h3>Feature 3: End-to-End Encryption Using Your Wallet Keys</h3><p>Many email services describe themselves as private, but the word “private” can mean different things.</p><p>For example, TLS encryption in transit protects a message while it travels between servers. Once the message reaches the server, that server can decrypt it.</p><p>That means the server can read it. The company can potentially be compelled to produce it. Someone with the right level of internal access could potentially view it.</p><p>Tmail approaches encryption differently.</p><p>When you send a message, your TON wallet’s cryptographic material encrypts the message on your device before transmission.</p><p>The encrypted version is what gets sent to storage.</p><p>When the recipient opens the message, their wallet generates the corresponding decryption material, and the message is decrypted locally on their device.</p><p>No server in this process holds the plaintext content.</p><p>The result is that nobody, including Tmail’s own developers, can access your email content.</p><p>This is not simply a privacy-policy statement. It follows from the architecture: decrypting the message requires wallet keys that Tmail never holds and has no mechanism to obtain.</p><h3>What the encryption experience feels like</h3><p>The user experience is designed to feel similar to standard email.</p><p>You connect your wallet once, then compose and receive messages without manually managing cryptographic keys or signing every individual message.</p><p>The encryption process happens underneath the interface while the service continues to behave like an email application.</p><p>The responsibility created by this model is the same one discussed in wallet sign-in: your recovery phrase is the root from which your encryption keys derive.</p><p>Lose it, and you permanently lose access to your encrypted mailbox.</p><h3>Feature 4: Decentralized Email Storage on IPFS and TON Storage</h3><p>Encryption addresses who can read your messages.</p><p>Storage architecture addresses another part of the problem: where your encrypted data lives and what happens if the storage provider becomes a point of failure.</p><p>Imagine storing encrypted messages on a single centralized company server.</p><p>Even with encryption, that server remains a central point of failure. A breach, legal order, acquisition, or policy change could influence how the data is accessed or managed.</p><p>Tmail stores messages and files on <strong>IPFS</strong> and <strong>TON Storage</strong>.</p><p>These are decentralized networks in which data is distributed across independent nodes rather than held in one location controlled by one organization.</p><p>The data stored on these networks is the ciphertext produced by the encryption system described above.</p><p>Even direct access to that stored data produces meaningless encrypted content without the wallet keys required to decrypt it.</p><p>There is no single server to compromise because the data is distributed.</p><p>There is no single organization holding readable content and encryption keys together because Tmail does not hold those keys.</p><p>There is also no corporate policy decision that can independently redefine how your data is stored because the storage infrastructure operates separately from any single company’s business.</p><h3>Tmail storage limits</h3><p><strong>Free plan:</strong> 500 MB</p><p><strong>Premium plan:</strong> 20 GB</p><p>Both plans use the same underlying architecture.</p><h3>Feature 5: Telegram-Native Tmail Integration</h3><p>Many new applications face the same adoption problem.</p><p>The product may be useful, but asking users to download another application, create another account, and learn another interface can be enough to stop adoption.</p><p>Tmail removes much of that friction by running natively inside <strong>Telegram</strong> as a Mini App and bot.</p><p>If you already use Telegram, you can open Tmail directly inside the application you already use.</p><p>There is no additional download.</p><p>There is no new standalone interface you need to learn from scratch.</p><p>There is no separate account system.</p><p>The TON wallet you may already use inside Telegram can serve as both your authentication mechanism and your encryption key.</p><h3>Tmail Premium payment options</h3><p>Premium can be paid for using:</p><ul><li>TON</li><li>USDT</li><li>GRAM</li><li>Telegram Stars</li></ul><p>This covers the main range of assets TON ecosystem users typically hold.</p><p>Transaction costs also benefit from TON’s architecture.</p><p>The Mail NFT naming system uses <strong>TON DNS</strong>, so your NFT-based email addresses exist within the same naming infrastructure used by other TON assets.</p><h3>Tmail’s broader roadmap</h3><p>The roadmap published on <strong>tmarket.icu</strong> describes a future in which documents, spreadsheets, domain names, games, and authentication can become available through the same unified interface built around Mail NFT identity.</p><p>That places the email service as the first layer of a broader personal productivity and identity stack.</p><p>In August, that roadmap also recorded <strong>TWallet Web at web.twallet.ae</strong> as the month’s product milestone.</p><h3>Feature 6: Web2 and Web3 Email in One Tmail Inbox</h3><p>A Web3 email service would have limited practical value if it could only communicate with other Web3 email users.</p><p>In August 2026, most email users still rely on Gmail, Outlook, and other traditional providers.</p><p>Tmail addresses that reality by giving users two functional addresses that route into the same inbox.</p><h3><a href="/service/mailto:yourname@tmail.ae">yourname@tmail.ae</a></h3><p>This works like a traditional email address.</p><p>Someone using Gmail can send an email to this address, receive a reply, and communicate normally without needing to know that the system behind the address is blockchain-based.</p><h3><a href="/service/mailto:yourname@tmail.ton">yourname@tmail.ton</a></h3><p>This address carries the full NFT ownership model, on-chain verification, and Web3 identity properties described earlier.</p><h3>One inbox, two worlds</h3><p>Both addresses deliver messages to the same inbox.</p><p>The encryption and decentralized-storage architecture also apply to the receiving side for both addresses.</p><p>For example, if a Gmail user sends a message to your <strong>tmail.ae</strong> address, your copy is stored on encrypted decentralized infrastructure and is accessible only through your wallet.</p><p>The sender’s copy continues to exist under the terms of their email provider.</p><p>Your copy exists under yours.</p><p>That dual-address structure is a major part of what makes Tmail usable as a primary email service rather than simply an experimental Web3 supplement.</p><p>You can begin using Tmail for new correspondence while existing relationships continue reaching you through <strong>tmail.ae</strong>.</p><p>The transition to a more fully on-chain inbox can therefore happen gradually without disrupting existing communication.</p><h3>Feature 7: Transfer by Tmail — Send Crypto Using an Email Address</h3><p>This is the first major August feature and one of the most practical additions to the Tmail ecosystem.</p><p>On <strong>5 August 2026</strong>, Tmail publicly announced the upcoming feature:</p><blockquote><em>“Next week, your Tmail name unlocks a new superpower. Send and receive digital assets using only your Tmail address. No wallet address required.”</em></blockquote><p>Source: <a href="https://x.com/tmail_ton/status/2084986580767289617">https://x.com/tmail_ton/status/2084986580767289617</a></p><p>On <strong>14 August 2026</strong>, Transfer by Tmail went live.</p><p>The official announcement stated:</p><blockquote><em>“You can now send assets using a Tmail address.”</em></blockquote><p>Bot: <a href="https://t.me/TheTmail?start=1">https://t.me/TheTmail?start=1</a></p><p>MiniApp: <a href="https://t.me/TheTmail/TSent">https://t.me/TheTmail/TSent</a></p><p>WebApp: <a href="https://app.tmail.ae/TSent">https://app.tmail.ae/TSent</a></p><p>Tmail also described the feature as:</p><blockquote><em>“This is just Step 1 of 4 in bringing transfers directly to your Tmail identity.”</em></blockquote><p>Source: <a href="https://x.com/tmail_ton/status/2088259301798232377">https://x.com/tmail_ton/status/2088259301798232377</a></p><h3>How the transfer works</h3><p>The product example published with the launch is:</p><p><a href="/service/mailto:your-name@tmail.ton"><strong>your-name@tmail.ton</strong></a><strong> → </strong><a href="/service/mailto:friend-name@tmail.ton"><strong>friend-name@tmail.ton</strong></a></p><p>The practical difference is straightforward.</p><p>Instead of copying a long hexadecimal wallet address, checking the first and last four characters, and hoping nothing went wrong in the middle, you simply enter a Tmail address.</p><p>The system resolves the Tmail identity to the underlying wallet.</p><p>Anyone who has sent crypto before knows how stressful copying and pasting a wallet address can be.</p><p>A readable email-style address is easier to remember, easier to recognize, and easier to verify than a 48-character wallet address.</p><h3>One Tmail NFT, multiple roles</h3><p>The Tmail NFT now serves three roles:</p><ol><li>Your email identity</li><li>The root of your encryption identity</li><li>Your human-readable payment address</li></ol><p>That means the Mail NFT is evolving from a simple inbox pointer into a broader identity layer.</p><p>NFT name ownership therefore becomes more valuable because the same name can function simultaneously as a communication address and a payment address.</p><h3>What is not yet confirmed</h3><p>Tmail officially described Transfer by Tmail as <strong>Step 1 of 4</strong>.</p><p>The August posts did not name Steps 3 and 4.</p><p>The official announcements also do not publish a complete list of supported assets or a full fee table.</p><p>Those details should be checked on the live TSent interface before treating any particular limit or supported-asset assumption as final.</p><h3>Feature 8: TWallet Web — Step 2 of the Transfer Program</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bn29jOp3A2r5afKBgpnUmw.png" /><figcaption>screenshot</figcaption></figure><p>On <strong>21 August 2026</strong>, Tmail expanded the transfer system beyond the inbox by launching a dedicated wallet product.</p><p>The official announcement said:</p><blockquote><em>“Step 2 is live: meet the web version of TWallet”</em></blockquote><p><a href="https://web.twallet.ae/">https://web.twallet.ae</a></p><p>It added:</p><blockquote><em>“You can now send tokens and collectibles using a Tmail address instead of a long wallet address.”</em></blockquote><p>And:</p><blockquote><em>“Next: TON Connect support and the TWallet Android app.”</em></blockquote><p>The post concluded:</p><blockquote><em>“One name. More possibilities.”</em></blockquote><p>Source: <a href="https://x.com/tmail_ton/status/2090786526196097362">https://x.com/tmail_ton/status/2090786526196097362</a></p><p>The Tmarket roadmap later recorded the same URL, <strong>web.twallet.ae</strong>, as the August 2026 TWallet Web milestone.</p><h3>What TWallet Web already supports</h3><p>According to the company, the web wallet can:</p><ul><li>Send tokens</li><li>Send collectibles</li><li>Use a Tmail name instead of a raw TON wallet address for the recipient</li></ul><h3>What is still coming</h3><p>The company described these as upcoming rather than live:</p><ul><li>TON Connect support on TWallet</li><li>TWallet Android app</li></ul><p>This is an important architectural shift.</p><p>Step 1 brought transfers into Tmail itself.</p><p>Step 2 introduced a named wallet domain under the same product family.</p><p>Tmail is therefore no longer only an email client that can also send assets. It now has a dedicated wallet surface where the destination format is the same Mail NFT identity that receives your email.</p><h3>Feature 9: TWallet in Telegram — Store, Swap, Send, and Receive Using a Tmail Name</h3><p>On <strong>28 August 2026</strong>, TWallet arrived inside Telegram, where Tmail users already spend their time.</p><p>The official announcement introduced:</p><blockquote><em>“Introducing TWallet in Telegram Mini App”</em></blockquote><p>It stated:</p><blockquote><em>“You can already store, send and receive coins and NFTs, swap assets, and make transfers using a TMail address instead of a long wallet address.”</em></blockquote><p>It also noted:</p><blockquote><em>“TON Connect support is coming soon.”</em></blockquote><p>Open TWallet: <a href="https://t.me/TWalletAppBot/app">https://t.me/TWalletAppBot/app</a></p><p>Source: <a href="https://x.com/tmail_ton/status/2093336939009130818">https://x.com/tmail_ton/status/2093336939009130818</a></p><h3>TWallet’s fullest public capability list</h3><p>The 28 August announcement provides the most complete public list of TWallet capabilities so far:</p><ul><li>Store coins</li><li>Store NFTs</li><li>Send and receive coins</li><li>Send and receive NFTs</li><li>Swap assets</li><li>Make transfers using a Tmail address instead of a long wallet address</li></ul><p>TON Connect was still described as coming soon.</p><p>The Android app announced on 21 August had also not yet been declared live as of 28 August.</p><h3>Why Telegram integration matters</h3><p>For people who already use Telegram, this is the most native financial surface Tmail has shipped so far.</p><p>The Tmail inbox Mini App and the TWallet Mini App now exist inside the same messenger.</p><p>The identifier used to access the mailbox can also be used to receive transfers.</p><h3>Feature 10: Gift Tmail Premium Using an Email Address</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZnyErmoZZnPkZP0Pc8ti9g.png" /><figcaption>screenshot</figcaption></figure><p>On <strong>25 August 2026</strong>, Tmarket introduced another feature built around the same identity model: Premium gifting.</p><p>The official announcement stated:</p><blockquote><em>“New feature on </em><a href="https://tmarket.icu/%E2%80%9D"><em>https://tmarket.icu/”</em></a></blockquote><p>It continued:</p><blockquote><em>“You can now gift a TMail Premium subscription to any account using its TMail email address.”</em></blockquote><p>Premium store:</p><p><a href="https://tmarket.icu/store/tmail/premium">https://tmarket.icu/store/tmail/premium</a></p><p>The announcement also said:</p><blockquote><em>“In the future, TMail Premium will extend across the entire ecosystem including projects that haven’t launched yet.”</em></blockquote><p>Source: <a href="https://x.com/tmail_ton/status/2092252328715858321">https://x.com/tmail_ton/status/2092252328715858321</a></p><p>The Telegram version of the announcement added two operational details that were shortened in the X post:</p><ul><li>No wallet address needed</li><li>The gift remains anonymous</li></ul><p>Before <strong>25 August</strong>, Premium was something a user purchased for their own connected account.</p><p>After <strong>25 August</strong>, Premium can be sent to a Tmail address in the same broader way that a Tmail name can be sent.</p><p>The second part of the official announcement is strategically significant: Premium is being designed as an ecosystem pass rather than simply an inbox upgrade.</p><p>The projects that may be included in that wider ecosystem have not been named in the August posts.</p><h3>Feature 11: Agentic Skills — Giving AI Agents Their Own Email Identities</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UzUdI4rr5K7NRCWhfnNyyg.png" /><figcaption>screenshot proof</figcaption></figure><p>Tmail’s agentic skills layer launched in <strong>July 2026</strong> and continued expanding through August at <strong>agents.tmail.ae</strong>.</p><p>The idea is to apply the Mail NFT identity model to non-human actors such as autonomous AI agents.</p><h3>The problem with AI agents using human email addresses</h3><p>Many AI agents that communicate through email currently use the email address of their human operator.</p><p>That creates several problems.</p><p>A recipient may not be able to tell whether a message came from a person or a machine.</p><p>Multiple agents using the same address can create attribution problems.</p><p>Emerging compliance frameworks around AI disclosure also have no straightforward way to distinguish and identify autonomous agents through conventional email infrastructure.</p><h3>Tmail’s approach to AI agent identities</h3><p>Tmail’s agentic skills layer allows an AI agent to hold its own <strong>Mail NFT at tmail.ton</strong> as a verifiable on-chain identity.</p><p>The NFT is tied to the wallet controlled by the agent’s operator.</p><p>A recipient can verify on the TON blockchain that the specific NFT exists and identify who holds it.</p><p>The agent’s correspondence history can therefore be attributed to its own identity rather than being mixed together with the operator’s personal email account.</p><p>The official July-to-August positioning says agents can:</p><ul><li>Send and receive Web2 and Web3 emails</li><li>Manage multiple addresses</li><li>Search emails</li><li>Maintain long conversation threads</li><li>Send files</li><li>Own their addresses as NFTs</li><li>Communicate with end-to-end encryption through a simple API</li></ul><p>Source: <a href="https://x.com/tmail_ton/status/2083203594409394454">https://x.com/tmail_ton/status/2083203594409394454</a></p><p>Docs and access:</p><p><a href="https://agents.tmail.ae/">https://agents.tmail.ae</a></p><h3>Agentic Skills vs. Tmail’s Premium AI Assistant</h3><p>This is different from the AI Assistant included with Premium plans.</p><p>The Premium AI Assistant helps human users with writing and inbox management.</p><p>The agentic skills layer creates email infrastructure for autonomous, non-human actors that can have their own accountable identities.</p><p>For builders working on multi-agent systems, customer communication agents, and research agents, the implications are practical.</p><p>Agents operating under their own Mail NFT identities can avoid attribution confusion.</p><p>Their on-chain identity can be verified.</p><p>Their correspondence can build an independent reputation instead of borrowing the reputation of their human operator.</p><p>The agentic skills layer is still under active development and its capabilities are expanding.</p><p>The foundational architecture is already in place.</p><p>It is not an August launch. Instead, it is the identity layer that now sits alongside the wallet and transfer work shipped during August.</p><h3>Tmail in August 2026: What Is Strong and What Is Still Being Built?</h3><p>Tmail’s architecture addresses a question traditional email has never been able to answer particularly well:</p><p><strong>Who actually controls your inbox?</strong></p><p>With the NFT ownership model, wallet authentication, wallet-key encryption, and decentralized storage working together, Tmail provides a clear answer.</p><p>You control the NFT.</p><p>You hold the keys.</p><p>The data is distributed.</p><p>The company operating the service does not have privileged access to your encrypted content.</p><h3>August expanded Tmail ownership into financial functionality</h3><p>August pushed this ownership model beyond communication.</p><p>The same identifier that receives your email can now receive tokens and collectibles.</p><p>TWallet Web and the TWallet Mini App give that identifier a dedicated wallet environment.</p><p>Gift Premium lets the same address receive a subscription without requiring a wallet address string.</p><h3>Where Tmail is still developing</h3><p>There are also important limitations that should remain clear.</p><p>Transfer by Tmail is officially <strong>Step 1 of 4</strong>.</p><p>TWallet TON Connect support and the Android application have been announced as upcoming rather than declared live.</p><p>The official posts do not yet publish complete TWallet custody details, a full asset list, or fee tables.</p><p>Those details need to be checked directly inside:</p><p><strong>web.twallet.ae</strong></p><p>and</p><p><strong>t.me/TWalletAppBot/app</strong></p><p>The wallet ecosystem is currently centered on TON.</p><p>A future multi-chain expansion could significantly increase the potential user base.</p><p>There is also a learning curve for people who have never used TON wallets or NFT-based identity systems.</p><p>Once the initial setup is complete, the ongoing experience is designed to be smooth, but the first-time setup requires users to understand concepts that traditional email users may never have encountered.</p><p>The agentic skills layer is architecturally powerful but still relatively early in terms of the range of available capabilities.</p><p>Its functionality is expected to expand as development continues.</p><h3>How to Get Started With Tmail in 6 Steps</h3><p>Getting started with Tmail follows a straightforward setup process.</p><h3>Step 1: Get a TON-Compatible Wallet</h3><p>Choose a TON-compatible wallet.</p><p><strong>Tonkeeper</strong> and <strong>Telegram Wallet</strong> are both straightforward options.</p><p>If you already use Telegram, Telegram Wallet is available directly inside the application.</p><h3>Step 2: Choose Your Tmail Name</h3><p>Browse available names through:</p><p><strong>tmarket.icu</strong></p><p>or secondary markets at:</p><p><strong>Getgems.io/tmail</strong></p><p>Purchase the name you want.</p><p>The NFT should arrive in your wallet within seconds.</p><h3>Step 3: Open Tmail and Connect Your Wallet</h3><p>Open Tmail through:</p><p><strong>tmail.ae</strong></p><p>or the Telegram Mini App.</p><p>Connect your wallet using <strong>TON Connect</strong>.</p><p>Your account is created automatically.</p><h3>Step 4: Secure Your Recovery Phrase</h3><p>Before sending or receiving anything important, write down your wallet recovery phrase and store it offline.</p><p>This is the most important security step in the entire setup process.</p><h3>Step 5: Choose Premium if You Need It</h3><p>Set up Premium if features such as:</p><ul><li>20 GB storage</li><li>AI assistance</li><li>Advanced spam filtering</li><li>Unlimited subdomains</li></ul><p>are relevant to your workflow.</p><p>Premium payments are accepted in:</p><ul><li>TON</li><li>USDT</li><li>GRAM</li><li>Telegram Stars</li></ul><p>You can also gift Premium to any Tmail address through:</p><p><a href="https://tmarket.icu/store/tmail/premium">https://tmarket.icu/store/tmail/premium</a></p><h3>Step 6: Start Using Transfers and TWallet</h3><p>Open transfers through:</p><p><a href="https://app.tmail.ae/TSent">https://app.tmail.ae/TSent</a></p><p>or:</p><p><a href="https://t.me/TheTmail/TSent">https://t.me/TheTmail/TSent</a></p><p>Open TWallet Web at:</p><p><a href="https://web.twallet.ae/">https://web.twallet.ae</a></p><p>Open TWallet inside Telegram at:</p><p><a href="https://t.me/TWalletAppBot/app">https://t.me/TWalletAppBot/app</a></p><p>Once your wallet is connected, both of your addresses are active immediately:</p><p><a href="/service/mailto:yourname@tmail.ae"><strong>yourname@tmail.ae</strong></a></p><p><a href="/service/mailto:yourname@tmail.ton"><strong>yourname@tmail.ton</strong></a></p><h3>Tmail Official Links and Product Access</h3><p><strong>Tmail:</strong><br><a href="https://tmail.ae/">https://tmail.ae</a></p><p><strong>NFT marketplace:</strong><br><a href="https://tmarket.icu/">https://tmarket.icu</a></p><p><strong>TWallet Web:</strong><br><a href="https://web.twallet.ae/">https://web.twallet.ae</a></p><p><strong>TWallet Mini App:</strong><br><a href="https://t.me/TWalletAppBot/app">https://t.me/TWalletAppBot/app</a></p><p><strong>Transfer Web:</strong><br><a href="https://app.tmail.ae/TSent">https://app.tmail.ae/TSent</a></p><p><strong>Agentic Skills:</strong><br><a href="https://agents.tmail.ae/">https://agents.tmail.ae</a></p><p><strong>Documentation:</strong><br><a href="https://docs.tmail.ae/">https://docs.tmail.ae</a></p><h3>Official Tmail August 2026 X Sources</h3><p><strong>Transfer announced:</strong><br><a href="https://x.com/tmail_ton/status/2084986580767289617">https://x.com/tmail_ton/status/2084986580767289617</a></p><p><strong>Transfer live, Step 1 of 4:</strong><br><a href="https://x.com/tmail_ton/status/2088259301798232377">https://x.com/tmail_ton/status/2088259301798232377</a></p><p><strong>TWallet Web, Step 2:</strong><br><a href="https://x.com/tmail_ton/status/2090786526196097362">https://x.com/tmail_ton/status/2090786526196097362</a></p><p><strong>Gift Premium:</strong><br><a href="https://x.com/tmail_ton/status/2092252328715858321">https://x.com/tmail_ton/status/2092252328715858321</a></p><p><strong>TWallet Mini App:</strong><br><a href="https://x.com/tmail_ton/status/2093336939009130818">https://x.com/tmail_ton/status/2093336939009130818</a></p><p><strong>Agent email layer:</strong><br><a href="https://x.com/tmail_ton/status/2083203594409394454">https://x.com/tmail_ton/status/2083203594409394454</a></p><h3>Final Note</h3><p>Always store your wallet recovery phrase offline before using any self-custodial service.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=251a4c54398c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why AI Agents Keep Forgetting Everything (And What Walrus Memory Does About It)]]></title>
            <link>https://emnaas.medium.com/why-ai-agents-keep-forgetting-everything-and-what-walrus-memory-does-about-it-4dd77b4b6849?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/4dd77b4b6849</guid>
            <category><![CDATA[cloud-storage]]></category>
            <category><![CDATA[memories]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[decentralized-storage]]></category>
            <category><![CDATA[walrus-memory]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Sun, 09 Aug 2026 19:15:41 GMT</pubDate>
            <atom:updated>2026-08-09T19:15:41.736Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*JjuUOliiRFgGk1epE5jpEg.png" /></figure><p>Imagine you are building a customer support agent. You spend a week training it on your product documentation. It handles tickets well. Users interact with it over three or four sessions. It starts to learn their preferences, their past issues, their communication style.</p><p>Then a session ends.</p><p>The next time the same user comes back, the agent has forgotten everything. It asks the same clarifying questions it already asked twice. It does not know what was tried last time. It does not know what failed. The user has to explain their situation from scratch, again, to a system that was supposedly learning about them.</p><p>This is not a failure of the underlying AI model. The model is fine. This is a failure of memory infrastructure. Specifically, it is what happens when memory is treated as a side effect of a session rather than a persistent, first-class asset that belongs to the user and travels with the agent.</p><p>Most current implementations of AI agent memory fall into one of two traps. Either memory exists only inside the active context window, which evaporates the moment the session ends, or memory is persisted inside a platform-specific database that the user does not control and that does not move between applications.</p><p>Both approaches produce the same outcome when you try to do anything sophisticated with an AI system: the agent hits reset at the worst possible moment.</p><h3>What Actually Breaks When Memory Is Locked</h3><p>The obvious failure is what the customer support example illustrates. The agent cannot continue where it left off.</p><p>But the less obvious failures are the ones that matter more for the developers building serious agent systems.</p><p>Consider a multi-agent workflow. You have a research agent that gathers information from external sources, a summarization agent that condenses findings, and a writing agent that produces reports. These three agents need to share state. The research agent needs to pass its findings to the summarization agent in a form the writing agent can later query. If memory is locked to a session or tied to a specific runtime, none of them can actually access a shared source of truth. Each agent operates in isolation while believing it is part of a coordinated system.</p><p>What breaks is not just continuity. What breaks is coordination. Agents make conflicting decisions because they each have incomplete context. Work gets duplicated because agents cannot tell what was already done. Complex jobs fail midway with no checkpoint, and there is no audit trail to trace why the agent acted the way it did.</p><p>There is also a trust problem. If agent memory lives inside a vendor’s proprietary database, you have no way to verify what the agent actually knows, what it has modified, or whether its stored context is intact. For applications where the accuracy of the agent’s knowledge matters, unverifiable memory is functionally unreliable memory.</p><p>The missing piece is an infrastructure layer that treats memory the way that storage infrastructure treats data: as a persistent, addressable, verifiable asset that belongs to whoever created it and can be accessed from wherever they choose.</p><h3>What Walrus Memory Is</h3><p>Walrus Memory launched in June 2026 as a portable memory layer specifically designed for AI agents. The core premise is direct: agent memory should be persistent across sessions, portable across applications and runtimes, verifiable by anyone, and controlled entirely by the user rather than by the platform delivering the service.</p><p>It is built on two infrastructure layers that handle separate responsibilities.</p><p>The first is Walrus, the decentralized storage protocol developed by Mysten Labs, the team behind the Sui network. Walrus stores data as immutable encrypted blobs distributed across a network of independent storage nodes using a bespoke erasure coding scheme called RedStuff. The architecture tolerates up to one-third of nodes being faulty or malicious while still allowing full data reconstruction. There is no single point of failure and no central operator holding your data. Walrus has stored over 450TB of data and currently prices storage at $0.023 per gigabyte per month.</p><p>The second layer is Sui, which provides the ownership and access control enforcement through smart contracts. The memwal::account Move module deployed on Sui mainnet defines who owns a Walrus Memory account, which keys are authorized to access it, and how permissions are granted to delegates. The smart contract does not store memory content. It manages identity, permissions, and access control, while the actual memory payloads live on Walrus in encrypted form.</p><p>These two layers together answer a question that most agent memory systems cannot: how do you make memory durable and verifiable without centralizing control?</p><h3>How the Storage Actually Works</h3><p>The flow for storing a memory, which Walrus Memory calls a “remember” operation, involves four steps that happen automatically when you call the SDK.</p><p>First, the relayer (the backend service that abstracts all of the Web3 complexity behind a standard REST API) generates a vector embedding from your plaintext content. The embedding is a numerical representation of the meaning of the memory, 1536 dimensions using text-embedding-3-small, and this is what makes semantic search possible later during recall.</p><p>Second, the relayer encrypts the plaintext content using Seal, Sui’s encryption framework. Only the account owner or their authorized delegates can decrypt the payload.</p><p>Third, the relayer uploads the encrypted payload to Walrus as a blob, attaching metadata including the namespace so that the restore process can discover the blob later.</p><p>Fourth, the vector embedding, along with the blob ID, owner address, and namespace, gets indexed in a PostgreSQL database with the pgvector extension. An HNSW index on the embedding column enables fast approximate nearest neighbor search when you need to recall memories.</p><p>Retrieving a memory inverts this flow. The relayer converts your query into a vector embedding, searches the database for the closest matching vectors using cosine distance, downloads the matching encrypted blobs from Walrus concurrently, decrypts each blob using Seal with the delegate key, and returns the plaintext results sorted by relevance. The most semantically relevant memories surface first.</p><p>The design means there are two layers with distinct responsibilities. Walrus holds the encrypted source of truth, durable and decentralized. PostgreSQL with pgvector holds the searchable index, fast and rebuildable. If the database is ever lost, the restore operation rediscovers blobs from Walrus by owner and namespace, re-embeds, and re-indexes them. The permanent record always lives on the decentralized layer.</p><h3>The Access Control Model and Why It Matters</h3><p>One property of Walrus Memory that separates it from conventional memory backends is programmable, onchain access control.</p><p>The smart contract maintains a list of authorized delegate keys for each account. Each delegate key is an Ed25519 public key that you register with your account. The relayer verifies every incoming request against this list before processing anything. Delegation is onchain, which means it is cryptographic, auditable, and cannot be forged by the relayer itself. Even if a relayer were compromised, it could not change who owns an account or create new delegate permissions without the account owner’s signature.</p><p>This enables a coordination pattern that is otherwise painful to implement: multiple agents operating under shared memory with scoped access. An agent can be granted access to a specific namespace without having access to the owner’s full memory. Access can be granted and revoked at any time. The agent’s permissions are enforced at the smart contract level, not by trusting the infrastructure provider’s promises.</p><p>Kostas Chalkias, Co-Founder and Chief Cryptographer at Mysten Labs, described why this matters: “Memory is one of the most critical bottlenecks in AI today. Most agent memory lives locked inside platforms. Walrus Memory changes this. It puts builders in control and lets agents move and collaborate across different services.”</p><h3>What You Can Actually Build</h3><p>The four memory operations in Walrus Memory (remember, recall, analyze, and ask) are designed as building blocks for different agent architectures.</p><p>Remember stores a memory with semantic understanding. Recall retrieves relevant memories using natural language queries, finding closest matches by meaning scoped to your memory space. Analyze extracts structured facts from text automatically, landing each fact as a separate memory for precise recall. Ask queries memories and returns an AI-generated answer with the relevant context attached, combining recall with LLM reasoning.</p><p>These primitives unlock a set of use cases that require persistent, cross-session memory.</p><p>A personal AI assistant that learns your tone, preferences, routines, and past decisions over months, rather than resetting every time you open a new chat. A customer support agent that knows what was tried, what failed, and what the customer actually needs, because it has read the prior session’s context from a portable memory store. A multi-step research workflow where a gathering agent, a synthesis agent, and a writing agent all read from and write to shared memory, coordinating on the same state without manual handoffs.</p><p>The AI Middleware integration is particularly relevant for teams building on Vercel AI SDK. It provides drop-in memory for AI conversations, automatically saving and recalling context around sessions without requiring manual plumbing of the storage operations.</p><h3>The Trust Model: Where to Place It</h3><p>Walrus Memory’s documentation is unusually honest about where trust lives and does not live in the system.</p><p>The relayer, which handles encryption, embedding, and decryption on behalf of the developer, sees plaintext content. That is a deliberate trade-off for developer experience. Web2 developers can integrate Walrus Memory without managing wallets, signing transactions, or understanding Sui directly because the relayer abstracts all of that. But the relayer operator can see data in transit.</p><p>Four deployment options address this at different trust levels. Using the managed relayer run by Walrus Foundation is convenient for prototyping. Self-hosting puts the infrastructure under your control. Running the relayer inside a Trusted Execution Environment (TEE) moves plaintext processing into an attested enclave. The manual client flow encrypts and embeds entirely on the client side, so the relayer only ever sees encrypted payloads and precomputed vectors and never touches plaintext.</p><p>What is enforced onchain and beyond the relayer’s ability to influence: ownership (only the owner’s private key controls the account), delegate authorization (the smart contract registers and verifies delegate keys), and access control (the contract determines who can act on an account). A compromised relayer cannot change ownership or forge permissions. The onchain enforcement is cryptographic and tamper-proof.</p><h3>The Developer Experience</h3><p>The TypeScript SDK wraps all Walrus Memory operations into a client your application calls directly.</p><pre>const memwal = MemWal.create({<br>  key: process.env.MEMWAL_PRIVATE_KEY,<br>  accountId: process.env.MEMWAL_ACCOUNT_ID,<br>  serverUrl: process.env.MEMWAL_SERVER_URL,<br>  namespace: &quot;my-app&quot;,<br>});</pre><pre>const job = await memwal.remember(&quot;User prefers dark mode and is located in Lagos.&quot;);<br>await memwal.waitForRememberJob(job.job_id);</pre><pre>const result = await memwal.recall({ query: &quot;What do we know about this user&#39;s preferences?&quot; });<br>console.log(result.results);</pre><p>A Python SDK is also available. The setup for using Walrus Memory with Claude Code, Cursor, Codex, or Gemini CLI is a single curl command that returns ready-to-use instructions.</p><p>The service is currently in beta and under active development.</p><h3>Why This Is the Right Architecture for the Agentic Future</h3><p>The pattern in every major era of software infrastructure has been the same. Someone takes something that was previously handled through bespoke, fragile, per-application implementations and turns it into standardized, portable, composable infrastructure.</p><p>Relational databases replaced custom file storage formats. Object storage replaced local disk management. Auth protocols replaced custom login implementations. The thing being standardized is rarely the most glamorous part of the stack, but the standardization always unlocks a generation of new applications that simply could not have been built when every team had to solve the problem from scratch.</p><p>Agent memory is at that inflection point now. Every serious AI application team has some version of the memory problem. Most are solving it with their own database, their own embedding pipeline, their own access control logic, and their own storage costs. None of those solutions are portable. None of them are verifiable. None of them give the user actual control.</p><p>Walrus Memory is an attempt to build that standardized infrastructure layer before the fragmentation becomes permanent. The combination of decentralized durable storage on Walrus, onchain ownership and access control on Sui, and a developer-friendly relayer API that hides the Web3 complexity behind standard HTTP calls is a coherent approach to a problem that genuinely requires all three parts to solve properly.</p><p>The specific insight that makes this architecture different from conventional approaches is the verifiability property. Because encrypted memory payloads live on Walrus as addressable, content-identified blobs, anyone can independently verify that a specific memory exists and has not been tampered with, without going through a centralized authority. For applications where the accuracy and integrity of agent knowledge matters, that verifiability is not a nice-to-have. It is the property that makes the memory actually trustworthy.</p><p>Sessions will still end. Context windows will still have limits. But the memory does not have to go with them. That is the problem Walrus Memory is built to solve, and understanding why it required this specific combination of decentralized storage, onchain access control, and vector-indexed retrieval helps clarify why the solutions that came before it were always only partial answers.</p><p><em>This article was written for the Walrus Memory Story Contest. Learn more at memory.walrus.xyz and follow @WalrusProtocol for updates.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4dd77b4b6849" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[YOUR EMAIL PROVIDER OWNS YOUR IDENTITY.TMAIL GIVE IT BACK TO YOU]]></title>
            <link>https://emnaas.medium.com/your-email-provider-owns-your-identity-tmail-give-it-back-to-you-666f96087087?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/666f96087087</guid>
            <category><![CDATA[blockchain]]></category>
            <category><![CDATA[email]]></category>
            <category><![CDATA[encryption]]></category>
            <category><![CDATA[nft]]></category>
            <category><![CDATA[top-email-platforms]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Tue, 28 Jul 2026 21:46:27 GMT</pubDate>
            <atom:updated>2026-07-28T21:46:27.582Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vhecTPnCK0z7ej7sGce8TQ.png" /></figure><p>Most people have not seriously questioned who owns their email address. It is such a routine piece of infrastructure that the question feels strange. But think about it for a moment.</p><p>The Gmail address you have used for ten years, the one that holds your receipts, your job applications, your personal conversations and your two-factor authentication recovery links, belongs to Google. Not you. Google decides whether you can keep it. Google can read every message stored on their servers. Google can comply with a legal request for your entire inbox without asking your permission. If Google suspends your account for any reason they define in a terms of service document that nobody reads, your address is gone along with everything tied to it.</p><p>This is not exceptional. It describes every major email provider that has ever existed. Outlook, Yahoo, Apple Mail, ProtonMail with its centralized key infrastructure. The model has always been the same. The company holds the infrastructure. The company makes the rules. The user accesses their account on the company’s terms.</p><p>Tmail is the first email service to challenge this architecture at every layer simultaneously. Not just the encryption. Not just the storage. Every layer, from the address itself to the authentication model to where the data lives to how the service integrates with the emerging AI agent economy.</p><p>This article explains each feature in full detail, what it is, how it works technically, and what it means practically for the person using it.</p><h3>NFT Email Addresses With True Ownership — Your Address as a Blockchain Asset</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TgcwsQQt_s4U8r6hFzh6-w.png" /><figcaption>SCREENSHOT OF USERNAMES</figcaption></figure><p>The most fundamental change Tmail makes to email is also the simplest to describe: your email address is an NFT.</p><p>This is not a metaphor. Your Tmail address, formatted as yourname@tmail.ton, is a non-fungible token registered on the TON blockchain using the TON DNS naming system. You purchase it on the official marketplace at tmarket.icu or on secondary markets at Getgems.io/tmail. After purchase, the NFT arrives in your TON wallet within seconds, exactly the same way any other NFT would.</p><p>What this changes is the ownership model in a way that has no precedent in email history.</p><p>When you own a Mail NFT, you can store it in any compatible TON wallet of your choosing. You can transfer it to another wallet, which transfers the inbox that corresponds to it. You can sell it on secondary markets at whatever price the market offers. You can gift it to someone else. You can accumulate multiple names and switch between them from a single Tmail interface with one click.</p><p>The TON blockchain enforces scarcity in a way that creates genuine market dynamics. Short names, common words, and names tied to recognizable identities carry measurable value because their supply is fixed. Secondary markets on Getgems.io/tmail reflect this with real transactions happening today, not in a speculative future.</p><p>The Free plan allows unlimited NFT mailboxes, meaning you can hold and manage as many separate addresses as you want without paying a subscription. Premium adds unlimited subdomains on each NFT name, formatted as your-nft-name.subdomain@tmail, which makes it practical to separate professional, personal, and project-specific correspondence under one identity without creating separate accounts.</p><p>The deeper point about ownership is this. Your email address has always been one of the most central identifiers in your digital life. Everything you sign up for asks for it. Everything that wants to reach you sends to it. Your access to other platforms frequently depends on it for password recovery. And for thirty years this critical piece of your digital infrastructure has existed entirely at the discretion of a company you have never met.</p><p>Making that address an on-chain NFT does not change how it looks or how it functions as an inbox. What it changes is who controls it and what you can do with it. The company providing the email service cannot revoke a TON NFT. The blockchain does not care whether the company remains in business. Your address and your inbox belong to whoever holds the token, permanently, on-chain.</p><h3>Sign In With Your Web3 Wallet — Authentication Without Passwords or KYC</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d5ox5faTGfgK_2rmUbuXqA.png" /><figcaption>I SIGHN IN AFTER CONNECTING MY TON WALLET</figcaption></figure><p>Tmail has no registration form. There is no password field, no email verification loop, and no KYC requirement of any kind.</p><p>You connect an active TON wallet through TON Connect and your account is created automatically. Authorization happens through a cryptographically secure wallet signature, the standard Web3 authentication method. Supported wallets include Tonkeeper, Telegram Wallet, and every other TON Connect compatible wallet.</p><p>The official documentation explains why this model is chosen:</p><p>“This login method is more reliable than passwords because you don’t need to remember or store passwords and the private key stays only with you. We don’t store your logins and passwords so they can’t be stolen from our servers. Authorization goes through cryptographically secure signature which is the Web3 standard.”</p><p>The security improvement over passwords is structural, not cosmetic. When you authenticate with a password, the server must verify it, which means the server must have some form of your credentials. Even with hashing, credential databases get breached. With wallet-based authentication, the server verifies a cryptographic signature. It never needs to store anything about your access credentials. There is nothing on the server side to steal.</p><p>The private key and recovery phrase remain entirely with you. Tmail’s own documentation is direct about the consequence:</p><p>“We don’t store your seed phrase, private keys, or other access credentials and no one, including us, can recover your wallet. If you lose access and haven’t saved your backup phrase, wallet recovery will be impossible.”</p><p>This is the correct design for a service that takes ownership and privacy seriously. The only way a service can guarantee it cannot read your messages or access your account is if it genuinely has no ability to do so. Wallet-based authentication is how that guarantee is made structurally real rather than a policy promise.</p><p>The practical requirement: back up your recovery phrase offline before depending on this account for anything important. This responsibility does not exist with Gmail because Gmail can reset your access. It exists with Tmail because the control is genuinely yours.</p><h3>End-to-End Encryption Secured by Your Own Secret Phrase — Real Privacy vs Promised Privacy</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5RxxDBN1kPtIX4fQasgi_w.png" /><figcaption>INFO GRAPHIC</figcaption></figure><p>Most email services that claim privacy are offering something weaker than what that word implies. TLS encryption in transit means your messages are encrypted while travelling between servers. The important omission is what happens when the message arrives. The server decrypts it. The server can read it. The server stores it in a form the company controls.</p><p>True end-to-end encryption means the message is encrypted on the sender’s device before it leaves and only decrypted on the recipient’s device when they open it. The servers in the middle never see anything but ciphertext, which is mathematical noise without the decryption key.</p><p>Tmail implements genuine end-to-end encryption by tying the encryption to something the user already has: their TON wallet and the secret phrase that generates it.</p><p>When you send an email on Tmail, your wallet’s cryptographic material encrypts the message on your device before transmission. What travels to storage is ciphertext. When your recipient opens the message, their wallet decrypts it locally. No server at any point in this chain ever touches plaintext content.</p><p>The official description is precise about what this means:</p><p>“Full Privacy: Only you and the recipient can access the data. No Third-Party Access: No one, including Tmail developers, can access your emails.”</p><p>The phrase “no one, including Tmail developers” is the key sentence. This is not a privacy policy. It is a technical constraint. Decryption requires the wallet keys, which Tmail never has and never stores. The company making the service cannot access your messages because the architecture prevents it, not because they promised to behave.</p><p>The recovery phrase is the root from which all cryptographic material in your Tmail account derives. The authentication keys, the encryption material, and the Mail NFT identity all trace back to that sequence of words. Losing the phrase means losing access to the corresponding encrypted mailbox permanently. Tmail cannot restore it. There is nothing to restore from, because the keys were never stored anywhere except in the wallet you controlled.</p><p>The experience on the surface is designed to feel like ordinary email. You connect your wallet once and then send and receive messages without signing individual transactions or managing keys manually. The complexity of the cryptographic system is hidden behind an interface that behaves like email. The difference from ordinary email is invisible until it matters, which is precisely when it counts.</p><h3>Decentralized Data Storage Powered by IPFS and TON Storage — No Central Server to Breach or Subpoena</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RQuCPamssSl8C8tyBUtoXQ.png" /><figcaption>INFOGRAPHIC</figcaption></figure><p>End-to-end encryption solves the content access problem but not the storage problem. If encrypted messages are stored on a single company’s centralized server, that server is still a target. A breach, a legal order, a policy change, or a company acquisition could change who controls the data even if it remains encrypted.</p><p>Tmail uses IPFS (InterPlanetary File System) and TON Storage for all message and file storage. Both are decentralized networks where data is distributed across independent nodes rather than stored in one physical location under one organization’s control.</p><p>The official statement on how this works together:</p><p>“All emails and files are securely stored in decentralized storages (IPFS and TON Storage) in encrypted form.”</p><p>The critical phrase is “in encrypted form.” What gets distributed across these decentralized networks is ciphertext, the encrypted output from the process described in Feature 3. Even if someone obtained the raw data from IPFS nodes, they would possess meaningless ciphertext without the wallet keys that can decrypt it.</p><p>The practical consequences of this architecture are significant for different threat scenarios.</p><p>There is no central server to breach because the data is distributed across independent nodes. A security incident affecting one node or one hosting provider does not expose the content because no single node holds a complete or decryptable copy.</p><p>There is no single company to compel with a legal order because the storage network is not owned by Tmail or any other single entity. A government request directed at Tmail cannot produce readable email content because Tmail does not hold the data or the keys.</p><p>There is no organizational decision that can result in your data being handled differently because the storage infrastructure operates independently of Tmail’s business decisions.</p><p>Storage capacity by plan:</p><p>The Free plan includes 500 MB of encrypted decentralized storage. The Premium plan includes 20 GB. Premium is available at 2.99 USDT per month or 35.88 USDT per year, with payment accepted in USDT, TON, GRAM, or Telegram Stars.</p><p>Both plans use the identical underlying architecture. The difference is capacity, not the security model.</p><h3>Seamless Integration With the TON Ecosystem and Crypto Wallets — Built Inside What You Already Use</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pnAhqyUz_8c2Hjyy-kHjGA.png" /><figcaption>INFO GRAPHIC</figcaption></figure><p>Tmail is not a separate application that happens to use blockchain infrastructure. It is natively built into the TON ecosystem and the Telegram environment that contains it.</p><p>The most immediate consequence of this is how users access the service. Tmail runs as a Telegram Mini App and as a Telegram bot. If you already use Telegram, you access Tmail directly inside the application you are already in without downloading a separate app, creating a new account, or switching between tools.</p><p>Authentication works through TON Connect with full support for every compatible TON wallet. Transaction fees benefit from TON’s architecture: the official positioning describes the network as offering high speed, low fees, and decentralization. Payments for Premium are accepted in TON, USDT, GRAM, and Telegram Stars, which reflects the range of assets TON ecosystem users typically hold.</p><p>The TON DNS naming system powers the @tmail.ton address format, meaning NFT email names live within the same naming infrastructure as other TON DNS assets. The Mail NFT itself is a standard TON NFT, which means it can be managed through any compatible TON wallet without any special tools or interfaces.</p><p>The roadmap published on tmarket.icu describes an expanding ecosystem where mail, documents, spreadsheets, domains, games, and authentication all become accessible from a unified interface built around the Mail NFT identity. This positions Tmail’s current features as the early layer of a broader personal productivity and identity stack rather than a standalone email application.</p><p>For Telegram users specifically, this integration means that the onboarding friction that stops most people from trying new tools essentially disappears. The wallet setup and the Mini App access are already part of the environment. The learning curve is the wallet backup, not the application.</p><h3>Web2 and Web3 Support in One Email Service — Staying Connected to Both Worlds</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5NcwzL3QsRZ-CQaXU4V8TQ.png" /><figcaption>SCREENSHOT HOW BOTH ARE</figcaption></figure><p>The practical argument for encrypted Web3 email collapses if it can only communicate with other encrypted Web3 email users. In mid-2026, the majority of email correspondents use Gmail, Outlook, and traditional providers. A service that cannot reach them is a sealed room, not an email service.</p><p>Tmail addresses this directly by giving every user two functional addresses in a single inbox.</p><p>The first address is yourname@tmail.ae, which is a standard Web2 format that functions identically to any traditional email address. Any person or service using Gmail, Outlook, Apple Mail, or any email client can send to this address and receive replies from it. The address looks and behaves like email that the recipient already understands.</p><p>The second address is yourname@tmail.ton, which is the on-chain NFT address described in Feature 1. This address functions within the Tmail ecosystem and the TON infrastructure, carrying all the ownership and encryption properties that the service provides.</p><p>Both addresses route to the same inbox. You see and manage all correspondence from one interface regardless of which address the sender used.</p><p>The security model applies to both addresses on the receiving side. When a Gmail contact sends a message to your tmail.ae address, your copy of that message is stored encrypted on decentralized infrastructure and is only accessible through your wallet. The sender’s copy lives on Google’s servers under Google’s terms. Your copy lives under yours.</p><p>The authentication model is also identical regardless of which address receives the message. You access everything through your TON wallet. There is no secondary password or alternative login for the Web2 address.</p><p>This design enables a gradual transition rather than a binary choice. You can begin using Tmail for new correspondence while maintaining existing relationships at your old address, without requiring your contacts to understand or adopt anything new. Over time, as more correspondents learn your Tmail address, the center of gravity shifts without any disruption to communication.</p><h3>Agentic Skills — What Happens When AI Agents Get Their Own Email Addresses</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2EMVL5bZJtsQig1Ai4lvyA.png" /><figcaption>INFO GRAPHIC</figcaption></figure><p>In July 2026, Tmail launched its agentic skills layer at agents.tmail.ae, described in official materials as a powerful automation layer for external agents.</p><p>To understand why this matters, consider what AI agents currently lack when they operate via email. Most agents today send messages using a human operator’s email address. The recipient has no way of knowing whether a human or a machine wrote the message. The on-chain record does not distinguish the agent’s actions from the operator’s. When multiple agents interact with a shared email address, attribution becomes impossible. Compliance frameworks that are developing around AI disclosure requirements have no mechanism to work with in this model.</p><p>Tmail’s agentic skills layer applies the same Mail NFT identity system described in Feature 1 to non-human actors. An AI agent can hold its own NFT email address at tmail.ton. The agent’s email identity is an on-chain asset tied to the wallet that controls the agent. The organization or individual responsible for the agent controls that wallet, making ownership and accountability verifiable from the blockchain record.</p><p>Because the agent’s identity is on-chain, several properties follow automatically. Anyone who receives email from agent-name@tmail.ton can verify on the TON blockchain that the corresponding NFT exists and who holds it. The agent’s correspondence history is attributable to that specific identity rather than shared with a human operator’s account. If the agent is transferred to a new team or replaced, the NFT can be transferred or a new identity established, and the historical record remains distinct.</p><p>The AI Assistant included in Premium is a separate feature for human users that helps with writing and reading personal email. The agentic skills layer is fundamentally different: it creates first-class email infrastructure for autonomous non-human actors rather than adding productivity tools to a human inbox.</p><p>The distinction between Premium’s AI Assistant and agents.tmail.ae reflects a clear design philosophy. Human users get tools to use their inboxes more effectively. AI agents get their own inboxes to operate with genuine identities.</p><p>The feature launched in July 2026 and capabilities will expand as agent use cases develop. The foundational architecture, Mail NFTs as verifiable on-chain identities for non-human actors operating through end-to-end encrypted, decentralized email infrastructure, is in place.</p><h3>The Honest Assessment</h3><p>Tmail’s architecture eliminates several attack vectors that traditional email cannot address. There is no central password database to steal. There is no server-side plaintext to read. There is no single company to compel for data access. There is no revocable account that a terms of service decision can eliminate.</p><p>The trade-offs are real and worth stating without minimizing them.</p><p>Losing your wallet recovery phrase means permanently losing access to your mailbox. This is not recoverable because recovery would require someone else to have your keys, which would defeat the entire security model. Storing the recovery phrase offline, in a secure location, is not optional when this is your primary email.</p><p>End-to-end encryption protects message content but not communication metadata. The fact that two Tmail addresses exchanged messages at a particular time may still be observable at the network level. This limitation applies across all current E2E email systems.</p><p>Users without any prior experience with TON wallets or blockchain transactions will face a learning curve at setup. The daily experience is designed to feel like standard email. The initial configuration requires familiarity with self-custodial wallet management that most email users have not needed before.</p><p>TON is currently the only supported blockchain. Multi-chain expansion is on the roadmap but not yet available.</p><h3>What Tmail Is When You See It All Together</h3><p>Most email privacy solutions patch one layer of the existing system. They add a privacy-respecting server, or they add an encryption layer while keeping the address under company control, or they add decentralized storage while keeping the authentication traditional.</p><p>Tmail restructures every layer at the same time.</p><p>The address is an on-chain asset you own, not a label a company grants you. Authentication happens through your wallet, not through credentials stored on a server. Message content is encrypted with your own cryptographic material before it leaves your device. Data is stored on decentralized networks in encrypted form. The service lives inside the TON and Telegram infrastructure you already use. Both the traditional email world and the on-chain identity world are supported from a single inbox. And now the same identity architecture is being extended to AI agents that need their own verifiable presence in the email ecosystem.</p><p>The result is an email service where ownership, security, and control are built into the architecture rather than promised by a company policy.</p><p>Your email address is a blockchain asset because you hold the token. Your messages are private because you hold the keys. Your data is secure because no single entity holds the storage. Your identity is verifiable because the blockchain provides the proof.</p><p>None of these properties depend on trusting Tmail. They depend on the TON blockchain, the cryptographic mechanisms, and the recovery phrase in your possession.</p><p>Try Tmail at <a href="https://ref.tmail.ae/dXqZDZ">tmail.ae</a></p><p>NFT marketplace: tmarket.icu</p><p><em>All feature details sourced from official Tmail documentation at tmail.ae and tmarket.icu as of July 2026. Always store your wallet recovery phrase offline before using any self-custodial service Thank you for reading.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=666f96087087" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is BOB Gateway Better Than THORChain? A 2026 Comparison With Chainflip & Garden Finance]]></title>
            <link>https://emnaas.medium.com/is-bob-gateway-better-than-thorchain-a-2026-comparison-with-chainflip-garden-finance-9a3a011392b6?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/9a3a011392b6</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[blockchain]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[swap]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Mon, 20 Jul 2026 16:36:25 GMT</pubDate>
            <atom:updated>2026-07-20T16:36:25.393Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3r5BYSsPJnsesNtmTkdT6g.png" /></figure><p>Something shifted in the native Bitcoin swap space in 2026.</p><p>Not gradually. Not through a slow decline in volume or user migration that you could explain away as normal market rotation. Something more specific happened that changed how serious Bitcoin holders and DeFi operators think about where they route their native BTC swaps.</p><p>THORChain — for years the unchallenged leader in native, non-custodial BTC swaps — was used as a laundering conduit in the $1.5 billion Bybit hack in February 2025. Then again in April 2026, the same infrastructure was used in the $293 million KelpDAO exploit.</p><p>THORChain was the routing layer in both cases, not the target. The protocol itself did not get hacked. But being the infrastructure that bad actors chose twice in fourteen months creates a specific kind of reputational weight that liquidity providers, institutional operators, and privacy-conscious traders do not dismiss easily.</p><p>The result: a serious, well-funded search for THORChain alternatives. Not from retail speculators looking for a quick arbitrage, but from exactly the users who move the most volume through native BTC swap protocols — and who have the most to lose from choosing the wrong venue.</p><p>This article answers that search with the data.</p><p>The direct answer: for the lowest fees on non-custodial native BTC to stablecoin swaps in 2026, <strong>BOB Gateway</strong> charges zero protocol fees and averages approximately 0.025% total all-in — the lowest documented cost of any non-custodial protocol currently operating this route. For chain coverage breadth, THORChain still leads. For execution certainty and liquidity depth on established pairs, Chainflip is competitive. For specific USDC routing efficiency, Garden Finance has moments.</p><p><em>Here is the full breakdown — not just the headline numbers, but the architecture behind each protocol and why it produces the outcomes it does.</em></p><h3>Why Protocol Architecture Determines Your Outcome (The Part Most Articles Skip)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zZDvQptUkswr_wdl-qvoXA.png" /><figcaption>INFOGRAPHIC</figcaption></figure><p>Before comparing platforms, it is worth understanding something that most comparison articles skip entirely: the architecture of a swap protocol determines its fee structure, its risk profile, and its execution certainty almost completely.</p><p>This sounds technical. It has a practical explanation.</p><p><strong>Pool-based protocols (THORChain model):</strong> You are swapping against a liquidity pool. The pool holds both BTC and RUNE. When you swap BTC for USDC, you are essentially selling BTC into the pool and buying RUNE out, then selling RUNE for USDC in another pool step. Every pool needs liquidity providers to deposit both assets. Those liquidity providers need to earn fees to stay profitable. The fee you pay is partly covering the LP yield. The slippage you experience is a function of how deep the pool is relative to your trade size.</p><p>This model works. THORChain has proven it at scale. But it embeds specific costs: LP yield requirements mean fees cannot go to zero, pool depth caps how efficiently large trades execute, and the RUNE intermediate step introduces a moment of volatility exposure mid-swap.</p><p><strong>AMM-based validator protocols (Chainflip model):</strong> Instead of individual LPs, Chainflip uses a 100-of-150 proof-of-stake validator network that manages vaults. Swaps route through USDC as the base pair — a design choice that eliminates the volatile intermediate asset problem THORChain has with RUNE. The validators stake FLIP tokens as security collateral and earn fees. The fee structure is more predictable because USDC as intermediate keeps pricing stable.</p><p><strong>Intent-based protocols (BOB Gateway model):</strong> This is architecturally distinct from both. You do not swap against a pool. You do not route through validators managing vaults. You publish an intent — a signed statement of exactly what you want: “I will send X BTC and I want to receive Y USDC.” Professional market makers (solvers) see this intent and compete to fill it. The solver who can source the best rate wins the order. Your Bitcoin and the solver’s USDC settle simultaneously through cross-chain proofs. No shared custody. No pool to drain. No intermediate asset.</p><p>The reason BOB Gateway can charge zero protocol fees is not charity. It is because the intent-based model does not require the same infrastructure overhead that pool-based or vault-based models need to sustain. The solvers earn a narrow spread on the execution. The protocol facilitates settlement rather than managing liquidity directly.</p><p>Understanding this changes how you read the fee comparison below. These are not just different prices for the same service. They are different services with different risk and execution profiles that happen to produce the same end result: native BTC converted to stablecoins.</p><h3>THORChain: The Established Standard With a Complicated 2025–2026</h3><p>THORChain launched in 2021 and became the dominant protocol for native, non-custodial BTC swaps. Its model: pooled BTC and RUNE liquidity, decentralized validator set, no KYC, broad chain coverage.</p><p>At its peak, THORChain processed hundreds of millions in daily volume. Today, its TVL sits at approximately $39.5 million — significantly below its operational high.</p><h4>What THORChain Does Well</h4><p><strong>Chain coverage is genuinely broad.</strong> BTC, ETH, BNB, AVAX, ATOM, DOGE, LTC, and a growing list of ERC-20 assets. If you need to swap BTC for an asset that other protocols do not support yet, THORChain is frequently the only non-custodial option available.</p><p><strong>Decentralization is real.</strong> No single entity controls the protocol. Validator selection is genuinely distributed. This is not a marketing claim — it is verifiable on-chain.</p><p><strong>No KYC required.</strong> Anyone with a BTC wallet can swap without identity verification. This is a meaningful property that THORChain has maintained consistently.</p><h4>The Fee Reality</h4><p>The February 2026 documented swap — 8.99 BTC ($618,000) for USDC — showed GemWallet charging 50 basis points as the interface fee, written directly into the transaction memo and permanently visible on-chain.</p><p>The fee structure: a protocol fee percentage plus a RUNE network fee of approximately $1–5 per swap. For large trades, the RUNE network fee is negligible. For smaller swaps, it can dominate the total cost.</p><p>For a $10,000 BTC-to-USDC swap using a standard 50 bps interface:</p><ul><li>Protocol + interface fee at 50 bps = $50</li><li>RUNE network fee = $2–5</li><li>Total approximate cost: $52–55</li></ul><h4>The On-Chain Transparency Question</h4><p>THORChain’s transparency is often cited as a feature. It is simultaneously a privacy consideration. Every detail of every swap — sending wallet, receiving wallet, exact amounts, timing, fees, memo data — is permanently visible on-chain. This served the documentation purpose in the February 2026 $618,000 swap analysis. It also means anyone watching the THORChain mempool and chain explorer can trace your transactions in real time, permanently.</p><p>For users who treat privacy as a factor in swap venue selection, this is relevant.</p><h4>The Hack Routing Context</h4><p>The $1.5B Bybit hack in February 2025 and the $293M KelpDAO hack in April 2026 both used THORChain as a laundering conduit. Again: THORChain was not hacked. The protocol processed these transactions as designed. The concern for certain operators is reputational and regulatory — being associated with infrastructure that processed proceeds from two of the largest crypto hacks in 2025–2026 creates friction with institutional counterparties and compliance teams.</p><p><strong>When THORChain is still the right choice:</strong> When you need chain or asset coverage that no other non-custodial protocol supports. The breadth advantage is real and meaningful for specific routes.</p><h3>Chainflip: The Most Technically Sophisticated Alternative</h3><p>Chainflip launched as a protocol with one specific design insight: THORChain’s use of RUNE as the intermediary asset in swaps introduces unnecessary volatility and execution risk. Chainflip chose USDC as its base pair instead.</p><p>Why does this matter? Because when your BTC routes through RUNE on THORChain, the RUNE price at the moment of swap determines your execution. RUNE is volatile. In a fast-moving market, the RUNE step can add execution uncertainty. When your BTC routes through USDC on Chainflip, the USDC step is stable. Pricing is predictable. The quote you receive reflects what you will actually get.</p><h4>The Validator Architecture</h4><p>Chainflip uses a 100-of-150 threshold signature scheme (TSS) for managing vaults — meaning 100 out of 150 validators must agree to authorize a vault operation. This is meaningfully different from THORChain’s validator model. It provides strong security guarantees while maintaining decentralization.</p><p>No wrapped tokens. No bridges. Validators hold native assets in their vaults and manage settlement directly.</p><h4>The Fee Structure in Detail</h4><ul><li><strong>Protocol fee:</strong> 0.10% flat on all swaps</li><li><strong>LP spread:</strong> 0.05–0.15% depending on pair liquidity depth</li><li><strong>Destination gas:</strong> Variable, approximately $2–10 depending on target chain</li></ul><p>For a $10,000 BTC-to-USDC swap:</p><ul><li>Protocol fee: $10</li><li>LP spread (0.10% mid-range): $10</li><li>Gas: $2–5</li><li>Total: approximately $22–25</li></ul><p>For a $100,000 swap:</p><ul><li>Protocol fee: $100</li><li>LP spread: $100</li><li>Gas: $2–5</li><li>Total: approximately $202–205</li></ul><p>The percentage-based cost remains constant. Gas becomes negligible at larger sizes.</p><h4>Volume and Liquidity Credentials</h4><p>Chainflip processed $6.39 billion in total volume across 2 million swaps since launch, as of February 2026. Monthly volume runs approximately $336 million. The BTC-to-USDT route alone accounted for $52.24 million in one 30-day window.</p><p>These are not small numbers. Chainflip has established real liquidity credibility, and the volume data makes its claim to competitive BTC swap routing credible rather than aspirational.</p><p>Chainflip Lending launched in February 2026, allowing users to borrow stablecoins against BTC collateral cross-chain — meaning Chainflip is expanding beyond swap execution into a broader financial product suite. This gives it a trajectory that resembles what BOB is building, though the specific products and mechanisms differ.</p><p><strong>When Chainflip is the right choice:</strong> When you need the deepest liquidity and most predictable execution on BTC-to-ETH, BTC-to-USDC, or BTC-to-SOL routes specifically. Chainflip’s established volume gives it real execution advantages on supported pairs.</p><p><strong>Where Chainflip falls short:</strong> At 0.21–0.30% total fees, it is meaningfully more expensive than BOB Gateway at scale. And its chain coverage, while growing, is narrower than THORChain.</p><h3>Garden Finance: The USDC Bridge Specialist</h3><p>Garden Finance operates with a specific focus that differs from the other protocols in this comparison: efficient USDC routing across EVM chains.</p><p>In a direct comparison of USDC bridging costs, Garden was identified as producing better output than Chainflip (approximately 0.57% loss on 100 USDC) and Symbiosis (approximately 0.71% loss) for the same transfer. Garden’s strength is on specific EVM-to-EVM USDC routes where it has optimized its liquidity and routing.</p><h4>What Garden Does Not Support</h4><p>This is the critical distinction for anyone doing a ThORChain alternatives search specifically for native BTC swaps:</p><p>Garden Finance does not support tokenized real-world assets (RWAs). Neither does THORChain.</p><p>BOB Gateway now supports tokenized gold — XAUT (Tether Gold) and PAXG (PAX Gold) — as part of its asset coverage. For a Bitcoin holder who wants exposure to both Bitcoin’s price appreciation and gold’s inflation hedge characteristics in a non-custodial structure, this is a genuinely differentiated capability that no other protocol in this comparison offers.</p><h4>The Integration Advantage BOB Has Built</h4><p>BOB has completed integrations in 2026 that expand its distribution significantly:</p><ul><li><strong>Socket integration:</strong> Connects BOB Gateway’s swap routing to Socket’s cross-chain aggregation infrastructure</li><li><strong>Velora integration:</strong> Extends coverage and routing options</li><li><strong>ShapeShift integration of BOB Gateway:</strong> ShapeShift’s user base now has access to BOB Gateway’s swap execution</li></ul><p>Garden Finance has not matched these integrations in breadth. THORChain’s ecosystem integrations are broad but established rather than expanding rapidly.</p><p><strong>When Garden Finance is the right choice:</strong> For small-to-medium USDC transfers within the EVM ecosystem where its specific route optimizations produce the best output. For users whose entire cross-chain activity lives within EVM and USDC routing, Garden competes well.</p><p><strong>Where Garden falls short:</strong> Native BTC swap support is not Garden’s primary design. RWA support does not exist. Integration breadth is narrower.</p><h3>BOB Gateway: The Zero-Fee Intent-Based Protocol</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H-JwLOfGfpltb214D0a1og.png" /><figcaption>INFOGRAPHIC</figcaption></figure><p>BOB Gateway launched native Bitcoin swaps in December 2025. It is built by the Bank of Bitcoin project, which raised $21 million from Coinbase Ventures, Castle Island Ventures, IOSG, and BTC miners.</p><p>The design philosophy starts with a specific observation: most Bitcoin holders who want to access DeFi or convert to stablecoins are forced to either use a CEX (with KYC and custody risk) or use wrapped BTC infrastructure (with custodian and peg risk). BOB Gateway was designed specifically to eliminate both compromises.</p><h4>How the Intent System Actually Works</h4><p>When you initiate a swap on BOB Gateway, you do not interact with a liquidity pool. The process is:</p><ol><li>You specify exactly what you want: “I will send 0.5 BTC and I want to receive USDC on Base”</li><li>BOB Gateway broadcasts this intent to its network of solvers</li><li>Solvers — professional market makers with capital — see the intent and respond with quotes</li><li>The solver offering the best rate is matched to your order</li><li>You send BTC to the solver’s designated Bitcoin address</li><li>Cross-chain proofs verify the incoming BTC transaction</li><li>The solver releases USDC to your EVM wallet atomically</li></ol><p>At no point does BOB hold your BTC. At no point does the solver hold your BTC in a shared pool. The settlement is atomic — both sides of the swap complete or neither does.</p><p>The fixed-rate quote you receive before confirming is the exact amount you will receive. Not an estimate. Not subject to slippage based on what happens in the pool between your quote and your execution. Exactly what the interface shows.</p><h4>The Fee Structure</h4><p>BOB Gateway charges <strong>zero protocol fees</strong>.</p><p>The total average cost of approximately 0.025% reflects the narrow spread that solvers earn on execution — this is the market-making margin, not a platform fee.</p><p>For a $10,000 BTC-to-USDC swap: approximately $2.50 total cost.<br> For a $100,000 swap: approximately $25 total cost.<br> For a $500,000 swap: approximately $125 total cost.</p><p>This is documented in BOB’s <a href="https://gobob.xyz/en/blog/btc-to-usdt-swap-fees-compared">fee comparison between BOB Gateway, THORChain, and NEAR Intents</a> and verified through <a href="https://bestquotes.gobob.xyz/">the live quote comparison tool at bestquotes.gobob.xyz</a>.</p><p>The practical statement from BOB’s own published data: “The bigger the trade, the bigger the difference.”</p><p>At $500,000, the difference between THORChain’s typical 0.50%+ and BOB Gateway’s 0.025% is approximately $2,375. That is not a marginal improvement. It is a material difference.</p><h4>ZeroConf: 10-Second Settlement</h4><p>BOB Gateway’s ZeroConf technology delivers destination assets in approximately 10 seconds using smart double-spend protection.</p><p>How this works without waiting for Bitcoin block confirmations (which take approximately 10 minutes): the solver accepts the risk of the incoming BTC transaction before confirmation, protected by BOB’s double-spend detection mechanism. The solver releases USDC immediately. If the BTC transaction is later detected as a double-spend attempt, the solver’s protection mechanism activates.</p><p>The practical result for the user: you send BTC, you have USDC in approximately 10 seconds.</p><p>For standard settlement without ZeroConf, the timeline is approximately 10 minutes — still competitive with THORChain’s 10–30 minute range and Chainflip’s 2–10 minutes on their supported routes.</p><h4>The RWA Differentiator</h4><p>BOB Gateway added support for tokenized gold — XAUT (Tether Gold) and PAXG (PAX Gold) — making it the only protocol in this comparison that allows a Bitcoin holder to swap directly into tokenized real-world asset exposure without intermediaries, custodians, or wrappers.</p><p>The significance: Bitcoin holders who want to reduce volatility exposure without leaving on-chain, non-custodial infrastructure now have a path to gold-backed assets without going through a CEX or trusting a custodian. This is new. THORChain does not offer it. Chainflip does not offer it. Garden Finance does not offer it.</p><p>For a detailed breakdown of how <a href="https://gobob.xyz/en/blog/swap-native-btc-onchain">native BTC swaps work on-chain through BOB Gateway</a>, BOB’s own documentation covers the full flow.</p><h3>The Honest Comparison Table</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*57XQHfB5eCp4BrZrnsxH_w.png" /></figure><h3>The Honest Caveat Section</h3><p><strong>Where THORChain still wins:</strong> Chain coverage. If your swap involves assets or chains not supported by BOB Gateway or Chainflip, THORChain’s breadth is the decisive factor. Nothing else in this comparison matches its asset universe.</p><p><strong>Where Chainflip wins:</strong> Liquidity depth on established pairs. $6.39 billion in processing history and $336M monthly volume represents real, mature liquidity. For execution certainty on BTC-ETH or BTC-SOL routes at large sizes where pool depth matters, Chainflip’s track record is meaningful.</p><p><strong>Where Garden Finance wins:</strong> Specific EVM-to-EVM USDC routing at small amounts. If you are routing less than $10,000 in USDC across EVM chains, Garden’s optimization for that specific corridor is worth checking.</p><p><strong>Where BOB Gateway is the clear answer:</strong> Fee optimization at any trade size, fixed-rate execution certainty, ZeroConf speed, and the only non-custodial RWA access from native BTC.</p><p><strong>What BOB Gateway cannot do yet:</strong> Assets and chains outside its current 11+ chain support. Very large positions where solver liquidity depth may set practical limits — check <a href="https://bestquotes.gobob.xyz/">bestquotes.gobob.xyz</a> before transacting to confirm available rates at your specific size.</p><h3>The Closing Point Worth Quoting</h3><blockquote><strong><em>The search for a THORChain alternative is really a search for the protocol that does what THORChain promises — non-custodial, no-KYC, native BTC swap execution — without the fee overhead, the slippage uncertainty, or the reputational weight that the hack routing incidents created.</em></strong></blockquote><blockquote><strong><em>On all three of those dimensions, BOB Gateway leads the comparison in 2026: zero protocol fees averaging 0.025% all-in, fixed-rate quotes that eliminate slippage by design, and a clean incident record backed by serious institutional funding.</em></strong></blockquote><blockquote><strong><em>Chainflip earns a strong second position on established pair liquidity depth. THORChain remains the only option for certain asset and chain combinations. Garden Finance serves its USDC niche well.</em></strong></blockquote><blockquote><strong><em>The best THORChain alternative for most native BTC swap use cases in 2026 is the one that charges the least, executes with the most certainty, and settles the fastest. That answer is BOB Gateway.</em></strong></blockquote><p><a href="https://medium.com/u/c0894405df33">BOB | Build On Bitcoin</a> | Live quotes: 🔗<a href="http://bestquotes.gobob.xyz/">LINK</a></p><p><em>Fee data sourced from Spark.money’s atomic swap protocol comparison (March 2026), Chainflip’s published burnonomics data (February 2026), BOB’s published fee documentation, and MEXC’s February 2026 THORChain transaction analysis. Always verify current rates. Not financial advice.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9a3a011392b6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Changelly vs ChangeNOW vs SwapSpace: Which Crypto Aggregator Has the Lowest Fees in 2026?]]></title>
            <link>https://emnaas.medium.com/changelly-vs-changenow-vs-swapspace-which-crypto-aggregator-has-the-lowest-fees-in-2026-ff8ec1c334bb?source=rss-e018c1600d49------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff8ec1c334bb</guid>
            <category><![CDATA[defi]]></category>
            <category><![CDATA[rocketx-exchange]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[kyc]]></category>
            <category><![CDATA[swap]]></category>
            <dc:creator><![CDATA[Emnaas]]></dc:creator>
            <pubDate>Sun, 19 Jul 2026 07:19:33 GMT</pubDate>
            <atom:updated>2026-07-19T07:19:33.010Z</atom:updated>
            <content:encoded><![CDATA[<h3>Changelly vs ChangeNOW vs SwapSpace: Which Crypto Aggregator Has the Lowest Fees in 2026? (<em>And Why the Answer Might Surprise You</em>)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FceKZ4S6_XXf6je2oQoURw.png" /></figure><p>If you have been comparing Changelly, ChangeNOW, and SwapSpace to find the lowest fees for your next crypto swap in 2026, this article gives you the honest breakdown.</p><p>The direct answer first: among the three platforms, fees are competitive and differences are marginal. ChangeNOW is generally the most predictable. SwapSpace sometimes wins on exotic routes when a partner has favorable liquidity. Changelly’s 0.25% headline fee looks attractive but the real cost is almost always higher once network fees, spread, and rate buffers are factored in.</p><p>But here is what the comparison reveals when you go deeper: all three platforms share the same structural limitation. None of them access both CEX and DEX liquidity simultaneously. That ceiling matters more than the difference between 0.25% and 0.30%.</p><p>The platform that solves this — <a href="https://medium.com/u/32d20cac9b9a">RocketX Exchange</a> — is not in the original comparison, but it consistently ends up as the answer when the goal is the best net rate across the full swap.</p><h3>Changelly: The Headline Fee vs The Real Cost</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DzEgsDpyhNnCKTnDuzewqw.png" /><figcaption>Infographic</figcaption></figure><p>Changelly has been in the instant swap market since 2015. The headline fee for floating-rate swaps is 0.25%. On paper, that sounds like one of the most competitive rates available.</p><p>The real cost is different.</p><p>According to Coin Bureau’s 2026 review, the actual cost of a Changelly swap depends on the final amount received after accounting for network fees, exchange rate movement, spread, fixed-rate pricing buffers, and any third-party provider fees. For fiat-to-crypto or crypto-to-fiat transactions, the real cost can be significantly higher than the 0.25% headline suggests.</p><p>Changelly can also trigger KYC verification at its discretion. Basic swaps may proceed without account creation, but AML checks on specific wallet addresses or transaction patterns can pause a swap unexpectedly. For users in the United States, service access may be restricted entirely.</p><p><strong>What Changelly does well:</strong> Wide asset support. Established reputation. Simple interface. Fixed-rate option available.</p><p><strong>Where it falls short:</strong> Real costs frequently exceed the headline fee. KYC risk on larger transactions. US restrictions limit availability.</p><p><strong>Chain coverage:</strong> Supports major blockchain networks but not the full multi-chain breadth of newer platforms.</p><h3>ChangeNOW: The Fastest Option With a KYC Asterisk</h3><p>ChangeNOW has built a strong reputation for speed. The platform supports over 1,400 cryptocurrencies with average swap times around two minutes for small to mid-size transactions. For users moving standard amounts across supported networks, it is consistently the fastest of the three.</p><p>The fee structure is competitive. Non-custodial by default, users retain wallet control throughout the swap.</p><p>The KYC question is where it gets nuanced. ChangeNOW applies an automated AML/risk prevention system that monitors transactions in real time. If the system detects a flag, the swap pauses and KYC verification through partner SumSub may be required. This has helped the platform stop fraudulent transactions. It also means “no KYC by default” is not the same as “no KYC guaranteed.”</p><p>For standard transactions below certain thresholds, ChangeNOW users typically complete swaps without verification requests. For larger amounts or transactions involving certain wallet histories, the experience may differ.</p><p><strong>What ChangeNOW does well:</strong> Fastest swap execution among the three. Competitive fees. Non-custodial architecture. 1,400+ assets.</p><p><strong>Where it falls short:</strong> KYC risk remains for flagged transactions. Chain coverage does not extend to the full multi-chain landscape of 2026.</p><h3>SwapSpace: The Aggregator That Shows Its Sources</h3><p>SwapSpace operates differently from Changelly and ChangeNOW. It is a crypto exchange aggregator that compares rates across partner providers, rather than processing swaps itself.</p><p>This design has one key advantage: transparency. SwapSpace shows users the KYC level of each partner exchange before they commit to a route, allowing informed choices about which provider handles the transaction.</p><p>The tradeoff is that the best rate on SwapSpace depends entirely on which partner has favorable liquidity at that moment. According to ChangeNOW’s own comparison analysis, SwapSpace is more variable — sometimes delivering a better deal when a partner has unusually favorable liquidity, but less predictable than ChangeNOW as a default.</p><p>SwapSpace itself does not require mandatory KYC. Partner exchanges it routes through may, depending on their own compliance requirements.</p><p><strong>What SwapSpace does well:</strong> Transparent sourcing. KYC visibility before committing. Can surface competitive rates on niche routes.</p><p><strong>Where it falls short:</strong> Rate variability. No direct control over swap execution quality. Coverage depends on partner network depth.</p><h3>Head-to-Head Comparison</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*11GUB2AtENXjozUoncUFDQ.png" /><figcaption>Infographic</figcaption></figure><h3>The Structural Problem All Three Share</h3><p>The comparison above shows meaningful differences between Changelly, ChangeNOW, and SwapSpace. What it also reveals is that all three operate on the same fundamental model: they route swaps through CEX-side liquidity only.</p><p>This matters for one specific reason. A CEX aggregator is limited to liquidity available on centralized exchanges. A DEX aggregator is limited to on-chain liquidity. When these two liquidity pools are not combined, users are always leaving potential rate improvement on the table — because the best rate for any given swap might exist on the CEX side, the DEX side, or a combination of both.</p><p>None of the three platforms in this comparison can access both simultaneously.</p><h3>Why RocketX Exchange Changes the Comparison</h3><p><a href="https://medium.com/u/32d20cac9b9a">RocketX Exchange</a> operates on a different model: hybrid CEX and DEX aggregation. Instead of routing through CEX liquidity only, RocketX simultaneously queries both centralized exchanges (including Binance, Kraken, OKX) and decentralized exchanges (including Uniswap, CowSwap) alongside bridge providers and other aggregators, then routes through whichever combination returns the best net price.</p><p>The practical results are documented in RocketX’s own rate comparison testing. According to<a href="https://www.rocketx.exchange/blog/rocketx-vs-cex-aggregator-comparison/"> RocketX’s comparison of CEX aggregator vs hybrid routing</a>, the price difference comes down to how many liquidity sources are checked before a price is quoted. A CEX-only platform prices against its own internal order book or partner CEX network. RocketX checks 500+ sources simultaneously.</p><p>Additional distinctions worth noting:</p><p><strong>Chain coverage:</strong> RocketX supports 200+ blockchain networks. Changelly, ChangeNOW, and SwapSpace focus primarily on major networks. For users needing to swap across newer or less common chains, the coverage gap is significant.</p><p><strong>KYC:</strong> RocketX does not require KYC. Non-custodial throughout — funds never pass through a RocketX-controlled wallet.</p><p><strong>Asset support:</strong> 20,000+ tokens across supported chains, including privacy coins and assets that CEX-only platforms frequently delist or restrict.</p><p><strong>Zero platform fees:</strong> RocketX charges zero platform fees on swaps under $100, with competitive pricing on larger amounts through the hybrid routing model.</p><h3>Honest Caveats</h3><p>No platform is right for every situation.</p><p>Changelly, ChangeNOW, and SwapSpace offer simpler interfaces that suit users who primarily swap between major assets on major chains. For straightforward BTC-to-ETH or USDT-to-SOL swaps in moderate amounts, any of the three will execute reasonably well.</p><p>RocketX has a more complex interface that reflects its broader capability. Users who are not comfortable navigating multi-chain settings may find the three simpler platforms easier to start with.</p><p><a href="https://medium.com/u/32d20cac9b9a">RocketX Exchange</a> hybrid model also means the quote comparison process involves more variables — comparing CEX routes alongside DEX routes requires understanding what each route involves in terms of custody during transit and execution time.</p><p>The 200+ chain coverage is a significant advantage for cross-chain power users. For users who only ever swap on Ethereum or BSC, this may not matter.</p><h3>The Bottom Line</h3><p>Comparing Changelly vs ChangeNOW vs SwapSpace in 2026 reveals genuine differences in speed, fee predictability, and KYC transparency. ChangeNOW wins on speed and predictability for standard swaps. SwapSpace wins on transparency about its partner network. Changelly’s headline fee is attractive but the real cost is consistently higher.</p><p>The more important finding is that all three operate within the same ceiling: CEX-only routing that leaves hybrid liquidity untapped.</p><p>For users who have outgrown that ceiling — who need cross-chain swaps across 200+ networks, access to both CEX and DEX liquidity simultaneously, zero KYC, and no platform fees on standard amounts — <a href="https://medium.com/u/32d20cac9b9a">RocketX Exchange</a> is the platform this comparison points toward as the answer to what all three are competing on.</p><p>The best swap aggregator is not necessarily the one with the lowest headline fee. It is the one that checks the most liquidity sources and delivers the best net amount to your wallet.</p><p><em>Data in this article draws from Coin Bureau’s 2026 Changelly review, ChangeNOW’s published comparison analyses, and RocketX Exchange’s documented rate comparison methodology. Always verify current rates and terms directly on each platform before executing a swap.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff8ec1c334bb" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>