How much does MiniMax M3.1 Flash cost?

MiniMax M3.1 Flash costs $0.075/M input tokens and $0.30/M output tokens, with separate rates for Cache Read at $0.0075/M tokens.

What is the context length of MiniMax M3.1 Flash?

MiniMax M3.1 Flash has a 1,048,576 token context window. It supports up to 131,072 completion tokens.

Does MiniMax M3.1 Flash support tool calling and structured outputs?

Yes. MiniMax M3.1 Flash accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

What inputs and outputs does MiniMax M3.1 Flash support?

MiniMax M3.1 Flash accepts text, images, and video as input and returns text.

Which providers serve MiniMax M3.1 Flash?

MiniMax M3.1 Flash is served by 11 providers on OpenRouter: Inferact vLLM, MiniMax, StreamLake, Fireworks, inference.net, DigitalOcean, Nebius Token Factory, SambaNova and 3 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

When was MiniMax M3.1 Flash released?

MiniMax M3.1 Flash was released on October 19, 2026.

The free launch period ended Oct 22, 2026. MiniMax M3.1 Flash is now served at list price ($0.15/M input, $0.60/M output, $0.015/M cached input).

MiniMax: MiniMax M3.1 Flash

minimax/minimax-m3.1-flash

Model weights

Compare

MiniMax-M3.1-Flash is a multimodal model from MiniMax. Compared with M3, it delivers significant improvements across multiple benchmarks, particularly in complex agentic tasks, coding, and cybersecurity. It is built for complex, multi-step co-work automation, understanding real codebases, fixing bugs, and building features, and visual understanding and creation, including frontend and presentation creation. It is an open-weight mixture-of-experts model with 23B active / 428B total parameters.

Modalities

In / out price

$0.075 / $0.30per 1M

Context

1.0M

Released

Oct 19, 2026

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput


	$0.15	$0.60	$0.015	0.17s	250 tok/s	99.42%
	$0.15	$0.60	$0.015	0.82s	133 tok/s	100.00%
	$0.15	$0.60	$0.015	0.70s	118 tok/s	99.95%
	$0.15	$0.60	$0.015	0.59s	105 tok/s	100.00%
	$0.075	$0.30	$0.0075	1.30s	100 tok/s	99.48%
	$0.15	$0.60	$0.015	0.39s	103 tok/s	95.51%
	$0.15	$0.60	$0.015	1.10s	143 tok/s	99.68%
10% off	$0.15$0.135	$0.60$0.54	$0.015$0.0135	1.19s	65 tok/s	99.94%
	$0.15	$0.60	$0.015	0.36s	64 tok/s	99.05%
	$0.15	$0.60	$0.015	1.41s	64 tok/s	98.53%
	$0.075	$0.30	$0.0075	0.55s	286 tok/s	100.00%
50% off	$0.30$0.15	$1.20$0.60	$0.03$0.015	0.97s	60 tok/s	100.00%

Input $0.15Output $0.60
250 tok/s99.42% uptime
Input $0.15Output $0.60
133 tok/s100.00% uptime
Input $0.15Output $0.60
118 tok/s99.95% uptime
Input $0.15Output $0.60
105 tok/s100.00% uptime
Input $0.075Output $0.30
100 tok/s99.48% uptime
Input $0.15Output $0.60
103 tok/s95.51% uptime
Input $0.15Output $0.60
143 tok/s99.68% uptime
Input $0.135Output $0.54
10% off
65 tok/s99.94% uptime
Input $0.15Output $0.60
64 tok/s99.05% uptime
Input $0.15Output $0.60
64 tok/s98.53% uptime
Input $0.075Output $0.30
286 tok/s100.00% uptime
Input $0.15Output $0.60
50% off
60 tok/s100.00% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.02252/M tokens

Output

$0.5068/M tokens

MiniMax M3.1 Flash — Price history

Price history

No effective price data for this selection

$0.0158	$0.300	87.7%	0.0%
$0.0206	$0.469	91.5%	0.2%
$0.0214	$0.600	95.3%	4.2%
$0.0219	$0.489	91.4%	84.8%
$0.0220	$0.599	94.8%	0.0%
$0.0240	$0.600	93.3%	2.4%
$0.0243	$0.599	93.1%	2.8%
$0.0271	$0.539	88.8%	0.3%
$0.0295	$0.599	89.3%	3.6%
$0.0323	$0.599	87.2%	0.1%
$0.0371	$0.599	83.6%	1.5%
$0.0696	$0.599	59.5%	0.0%

Effective in $0.0158Effective out $0.300
Token share
0.0%
Effective in $0.0206Effective out $0.469
Token share
0.2%
Effective in $0.0214Effective out $0.600
Token share
4.2%
Effective in $0.0219Effective out $0.489
Token share
84.8%
Effective in $0.0220Effective out $0.599
Token share
0.0%
Effective in $0.0240Effective out $0.600
Token share
2.4%
Effective in $0.0243Effective out $0.599
Token share
2.8%
Effective in $0.0271Effective out $0.539
Token share
0.3%
Effective in $0.0295Effective out $0.599
Token share
3.6%
Effective in $0.0323Effective out $0.599
Token share
0.1%
Effective in $0.0371Effective out $0.599
Token share
1.5%
Effective in $0.0696Effective out $0.599
Token share
0.0%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

286tok/s

P50, best across providers

Latency

0.17s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchRelace86.5%74.7%Novita89.3%70.9%Nebius Token Factory87.5%72.6%inference.net87.0%73.0%Together87.6%72.0%inference.net (US)85.0%67.3%

+11 more providers

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)

100.00%

Availability (3d)

99.75%

Availability over the last 3 days

Last 72 hours

Availability 99.75%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.46%

Without Routing

98.63%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for MiniMax: MiniMax M3.1 Flash (Artificial Analysis and Vals AI)
Source	Benchmark	Score
Artificial Analysis	MiniMax-M3.1-Flash Intelligence Index	29.1
Artificial Analysis	MiniMax-M3.1-Flash HLE	48.0%
Artificial Analysis	MiniMax-M3.1-Flash AA-LCR	79.7%
Artificial Analysis	MiniMax-M3.1-Flash GDPval-AA	48.2%
Artificial Analysis	MiniMax-M3.1-Flash CritPt	22.3%
Artificial Analysis	MiniMax-M3.1-Flash SciCode	53.4%
Artificial Analysis	MiniMax-M3.1-Flash Terminal-Bench 4.0	25.8%
Artificial Analysis	MiniMax-M3.1-Flash AA-Omniscience Accuracy	45.1%
Artificial Analysis	MiniMax-M3.1-Flash AA-Omniscience Non-Hallucination Rate	6.3%
Vals AI	MiniMax-M3.1-Flash Access to Justice Benchmark	73.3%
Vals AI	MiniMax-M3.1-Flash EMB v1.1	67.8%
Vals AI	MiniMax-M3.1-Flash Finance Agent v2.1	58.8%
Vals AI	MiniMax-M3.1-Flash Vals Finance Index	55.9%
Vals AI	MiniMax-M3.1-Flash HLAB-Vals-1.1	7.5%
Vals AI	MiniMax-M3.1-Flash Vals Legal Index	67.1%
Vals AI	MiniMax-M3.1-Flash Legal Research Bench	34.1%
Vals AI	MiniMax-M3.1-Flash SAGE	85.3%
Vals AI	MiniMax-M3.1-Flash Tax Agent Bench	32.1%
Vals AI	MiniMax-M3.1-Flash Terminal-Bench 4.0	25.3%
Vals AI	MiniMax-M3.1-Flash Vals Index	53.8%
Vals AI	MiniMax-M3.1-Flash Vibe Code Bench v1.1	73.9%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

Hermes Agent

Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.

1.97Ttokens

OpenClaw

OpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.

1.03Ttokens

Kilo Code

Kilo Code is an open-source AI coding agent that works across VS Code, JetBrains, and CLI to help developers ship code faster with agentic workflows.

681Btokens

Mira is the leading AI agent inside Telegram

Mira is a Telegram-native AI assistant.

388Btokens

There are many coding agents, but this one is yours.

151Btokens

Rest of OpenRouter

302B

Top 5 Apps

124B

Not currently available through EU in-region routing. In-region routing docs

More models from MiniMax

MiniMax H3 Image Highspeed

MiniMax H3 Image Highspeed is the faster version of MiniMax's H3 Image model. It takes the same inputs and limits as MiniMax H3 Image: it creates images from text prompts and edits or composes images from 1 to 16 reference images, with output up to 20 megapixels and aspect ratios from 1:3 to 3:1.

Image$3 / $7.60

MiniMax H3 Image

MiniMax H3 Image is MiniMax's image generation and editing model. It creates images from text prompts and edits or composes images from 1 to 16 reference images, with output up to 20 megapixels and aspect ratios from 1:3 to 3:1.

Image$1.50 / $3.80

H3 Max

MiniMax H3 Max is a video-generation model from MiniMax, jointly released with fal.ai. Derived through additional training from MiniMax H3, it is designed for faster text-to-video and image-to-video generation with controlled first-frame or last-frame keyframes.

Videofrom $0.05/second

MiniMax H3 is a lightweight, open-weights video generation model from MiniMax. It is designed for precise multimodal editing and controlled content generation, including instruction-guided edits, text and brand rendering, and video-to-video motion transfer.

The model is suited for commercial creative workflows across advertising, e-commerce, gaming, and interface design, with native audiovisual output for reference-driven generation.

Videofrom $0.13/second

Speech 2.8 HD

MiniMax Speech 2.8 HD is a text-to-speech model from MiniMax. It is suited for applications that generate spoken audio from text and accepts arbitrary MiniMax voice IDs.

Speech$100/M characters

Speech 2.8 Turbo

MiniMax Speech 2.8 Turbo is a text-to-speech model from MiniMax. It is suited for applications that generate spoken audio from text and accepts arbitrary MiniMax voice IDs.

Speech$60/M characters

MiniMax M3

MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full attention with KV-block selection to cut per-token compute at long context — roughly 1/20 the cost of the previous generation at 1M tokens, with substantially faster prefill and decode while retaining quality across most tasks.

Trained as a native multimodal model on interleaved data and tuned for multi-turn, production-like collaboration via an interactive user-simulator framework, the model is oriented toward sustained, multi-step tasks rather than single-turn execution.

Text1.0M context$0.23 / $0.96

MiniMax M3 (batch)

Text524K context$0.224 / $0.88

Hailuo 2.3

Hailuo 2.3 is a video generation model from MiniMax. It accepts text prompts and reference images as input and generates video output, supporting both text-to-video and image-to-video workflows. It is suited for creative content production, cinematic scene generation, and character animation, with a focus on realistic motion and expressive character rendering.

Video$0.0817/second

MiniMax M2.7

MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Built to actively participate in its own evolution, M2.7 integrates advanced agentic capabilities through multi-agent collaboration, enabling it to plan, execute, and refine complex tasks across dynamic environments.

Trained for production-grade performance, M2.7 handles workflows such as live debugging, root cause analysis, financial modeling, and full document generation across Word, Excel, and PowerPoint. It delivers strong results on benchmarks including 56.2% on SWE-Pro and 57.0% on Terminal Bench 2, while achieving a 1495 ELO on GDPval-AA, setting a new standard for multi-agent systems operating in real-world digital workflows.

Text197K context$0.21 / $0.84

MiniMax M2.5

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1 to extend into general office work, reaching fluency in generating and operating Word, Excel, and Powerpoint files, context switching between diverse software environments, and working across different agent and human teams. Scoring 80.2% on SWE-Bench Verified, 51.3% on Multi-SWE-Bench, and 76.3% on BrowseComp, M2.5 is also more token efficient than previous generations, having been trained to optimize its actions and output through planning.

Text198K context$0.27 / $0.95

MiniMax M2-her

MiniMax M2-her is a dialogue-first large language model built for immersive roleplay, character-driven chat, and expressive multi-turn conversations. Designed to stay consistent in tone and personality, it supports rich message roles (user_system, group, sample_message_user, sample_message_ai) and can learn from example dialogue to better match the style and pacing of your scenario, making it a strong choice for storytelling, companions, and conversational experiences where natural flow and vivid interaction matter most.

Text66K context

MiniMax M2.1

MiniMax-M2.1 is a lightweight, state-of-the-art large language model optimized for coding, agentic workflows, and modern application development. With only 10 billion activated parameters, it delivers a major jump in real-world capability while maintaining exceptional latency, scalability, and cost efficiency.

Compared to its predecessor, M2.1 delivers cleaner, more concise outputs and faster perceived response times. It shows leading multilingual coding performance across major systems and application languages, achieving 49.4% on Multi-SWE-Bench and 72.5% on SWE-Bench Multilingual, and serves as a versatile agent “brain” for IDEs, coding tools, and general-purpose assistance.

To avoid degrading this model's performance, MiniMax highly recommends preserving reasoning between turns. Learn more about using reasoning_details to pass back reasoning in our docs.

Text205K context

MiniMax M2

MiniMax-M2 is a compact, high-efficiency large language model optimized for end-to-end coding and agentic workflows. With 10 billion activated parameters (230 billion total), it delivers near-frontier intelligence across general reasoning, tool use, and multi-step task execution while maintaining low latency and deployment efficiency.

The model excels in code generation, multi-file editing, compile-run-fix loops, and test-validated repair, showing strong results on SWE-Bench Verified, Multi-SWE-Bench, and Terminal-Bench. It also performs competitively in agentic evaluations such as BrowseComp and GAIA, effectively handling long-horizon planning, retrieval, and recovery from execution errors.

Benchmarked by Artificial Analysis, MiniMax-M2 ranks among the top open-source models for composite intelligence, spanning mathematics, science, and instruction-following. Its small activation footprint enables fast inference, high concurrency, and improved unit economics, making it well-suited for large-scale agents, developer assistants, and reasoning-driven applications that require responsiveness and cost efficiency.

To avoid degrading this model's performance, MiniMax highly recommends preserving reasoning between turns. Learn more about using reasoning_details to pass back reasoning in our docs.

Text205K context

MiniMax M1

MiniMax-M1 is a large-scale, open-weight reasoning model designed for extended context and high-efficiency inference. It leverages a hybrid Mixture-of-Experts (MoE) architecture paired with a custom "lightning attention" mechanism, allowing it to process long sequences—up to 1 million tokens—while maintaining competitive FLOP efficiency. With 456 billion total parameters and 45.9B active per token, this variant is optimized for complex, multi-step reasoning tasks.

Trained via a custom reinforcement learning pipeline (CISPO), M1 excels in long-context understanding, software engineering, agentic tool use, and mathematical reasoning. Benchmarks show strong performance across FullStackBench, SWE-bench, MATH, GPQA, and TAU-Bench, often outperforming other open models like DeepSeek R1 and Qwen3-235B.

Text1M context

MiniMax-01

MiniMax-01 is a combines MiniMax-Text-01 for text generation and MiniMax-VL-01 for image understanding. It has 456 billion parameters, with 45.9 billion parameters activated per inference, and can handle a context of up to 4 million tokens.

The text model adopts a hybrid architecture that combines Lightning Attention, Softmax Attention, and Mixture-of-Experts (MoE). The image model adopts the “ViT-MLP-LLM” framework and is trained on top of the text model.

To read more about the release, see: https://www.minimaxi.com/en/news/minimax-01-series-2

Text1M context

MiniMax: MiniMax M3.1 Flash

Providers