Nomily AI·2d agoRun your meeting transcriptions at home: faster-whisper + Ollama + Nomily (BYOK)Your recordings stay on your LAN. Your key, your ASR, your AI — no forced Nomily subscription.
Himanshutiwari·Nov 1Deepgram vs faster-whisper: Testing Hinglish Speech Recognition for Real-Time Voice AgentsVoice AI is growing rapidly, but building a voice agent that understands real-world conversations is still challenging.
InPython in Plain EnglishbyZhang Zhaoyu·Oct 30Top Trusting ASR Timestamps: Build a Transcript-Aware Audio Clip QA Gate with Python and FFmpegA reproducible pipeline for turning approximate speech timestamps into validated, traceable audio clips
Reda Malki·Oct 27Teaching Whisper Moroccan Darija: 6.26% WER with LoRA, Benchmarked and Shipped to WhatsApp
PielouNW·Oct 26How to build an offline transcription app with WhisperAt NobodyWho, we’ve spent the last months shipping new features, like Speech-to-Text or Voice Activity Detection, to complete our local AI…
Kiwi Tools·Oct 24How to transcribe podcasts, audio and video to text and SRT subtitles with Whisper (no GPU needed)Disclosure: We’re Kiwi Tools, and we built and sell the tool this tutorial uses: an Apify Actor called Audio & Video Transcriber. It…A response icon1A response icon1
WrenTools·Oct 21Transcribe audio and video files to text and SRT subtitles with about 50 lines of PythonTurn a list of audio or video links into .txt transcripts and .srt subtitles with Whisper, and a hard cost cap
asmodeus·Oct 21The dev set lied: fine-tuning Whisper for Javanese on a free GPUWhat ten test submissions to Mozilla’s Lost in Transcription challenge taught me about which numbers to trust.
InPython in Plain EnglishbyZhang Zhaoyu·Oct 21Stop Shipping Broken AI Transcripts: Build a Python QA Gate for Whisper Hallucinations, Repetition…
Varun Nuthalapati·Oct 15Speech-to-Text in 16.9 MBHow Cactus’s Whistle fits a seven-language transcriber into one tiny CPU-only file — and what its benchmarks do and don’t prove
