Kyutai Moshi vs Threadfork

A side-by-side look at pricing, capabilities, pros, cons, and our editorial scores.

	Kyutai Moshi Audio	Threadfork Audio
Tagline	Open-source, full-duplex speech-to-speech foundation model with sub-200ms latency	Private, local-first AI meeting notetaker for macOS
Category	Audio	Audio
Pricing	Free· Free and open source. Models under CC-BY 4.0, code under MIT (Python) / Apache 2.0 (Rust). Self-hosted only — you pay your own compute (24GB+ GPU for PyTorch, or Apple Silicon via MLX).	Paid· Pro: $39
Model	Moshi (7B-class speech-text foundation model) + Mimi neural audio codec, in-house by Kyutai	On-device local LLMs running on Apple Silicon (specific model family undisclosed)
Editorial score	—	—
Use cases	Real-time voice assistant prototypesResearch on full-duplex spoken dialogueOn-device voice interaction on Apple Silicon via MLXLow-latency conversational agents behind WebSocketNeural audio codec experimentation with MimiSelf-hosted voice interface for privacy-sensitive appsSpeech tokenizer for downstream audio LLM trainingInterruptible in-car or wearable voice UX	Confidential client meeting transcriptionSales call notes and follow-upsConsulting engagement documentationLegal intake and interview recordingTherapy or coaching session notesFounder investor-call archiveCross-meeting commitment trackingSemantic search over past conversationsOffline meeting capture while travelingEntity timelines for accounts and stakeholders
Pros	Truly full-duplex — handles interruptions, overlap and back-channels rather than rigid turn-taking Sub-200ms practical latency on a single L4 GPU, well below third-party voice APIs Fully open weights (CC-BY 4.0) plus MIT/Apache code — self-host with no per-minute billing Ships with Mimi, a streaming neural audio codec that beats SpeechTokenizer and SemantiCodec Multiple inference backends: PyTorch for research, Rust/Candle for production, MLX for on-device Mac/iPhone Inner-monologue text prediction gives you a transcript alongside the audio stream for free	Fully local, on-device transcription and summarization — no audio leaves the Mac Works offline, useful for travel and air-gapped environments Captures both system audio and mic, so it covers Zoom/Meet/Teams and in-person meetings from the same app Extracts structured commitments with owners and deadlines, not just a raw summary Cross-meeting entity timelines for people, companies and topics build up a real institutional memory Semantic search across the full transcript archive Privacy posture is a genuine differentiator versus cloud notetakers like Otter, Fireflies, Fathom and Granola
Cons	English-only voices at launch — no multilingual support out of the box Knowledge and reasoning quality trail top text LLMs; it's a 7B-class model, not GPT-4o Voice Requires a 24GB+ GPU for the reference PyTorch build; on-device is only viable via MLX on Apple Silicon No hosted API or SaaS tier — you own the ops, scaling and safety filtering Only two fixed synthetic voices (Moshiko/Moshika); no voice cloning or speaker conditioning in the release	macOS only — no Windows, Linux, iOS, web or mobile companion app Requires Apple Silicon; older Intel Macs are excluded $39/month is noticeably more expensive than most cloud notetakers, and there is no cheaper individual tier 3-hour cap per recording will bite long workshops, all-hands and depositions Local models generally trail frontier cloud LLMs on nuanced summarization and multilingual accuracy No public detail on which underlying local models are used, which complicates evaluation for compliance reviews
Website	kyutai.org	www.threadfork.com

Pick Kyutai Moshi if

✅ Truly full-duplex — handles interruptions, overlap and back-channels rather than rigid turn-taking
✅ Sub-200ms practical latency on a single L4 GPU, well below third-party voice APIs
✅ Fully open weights (CC-BY 4.0) plus MIT/Apache code — self-host with no per-minute billing
✅ Ships with Mimi, a streaming neural audio codec that beats SpeechTokenizer and SemantiCodec

Pick Threadfork if

✅ Fully local, on-device transcription and summarization — no audio leaves the Mac
✅ Works offline, useful for travel and air-gapped environments
✅ Captures both system audio and mic, so it covers Zoom/Meet/Teams and in-person meetings from the same app
✅ Extracts structured commitments with owners and deadlines, not just a raw summary

Compare a different pair →