Kyutai (Moshi / Unmute) vs Cartesia vs Fish Audio vs Whisper API
A side-by-side look at scores, pricing and features — with RECATOOLS' ASEAN-aware verdict for each.
|
|
Cartesia
Low-latency speech models built on state space models, for voice agent...
Visit
|
Fish Audio
Expressive TTS and voice cloning in 80+ languages, with open weights
Visit
|
Whisper API
OpenAI's open speech-to-text model; the hosted API retires in 2027
Visit
|
|
|---|---|---|---|---|
| RECATOOLS Score | 7 / 10 | 7.6 / 10 | 7.4 / 10 | 8.3 / 10 |
| Capability | ||||
| Value for money | ||||
| Ease of use | ||||
| ASEAN readiness | ||||
| API quality | ||||
| Pricing | Open Source | Freemium | Freemium | Freemium |
| Free tier | All code (MIT/Apache-2.0) and weights (CC-BY 4.0) free to self-host; free web demos at unmute.sh and moshi-chat.kyutai.org | Free: 20K credits a month (about 27 minutes of Sonic speech) plus $1 a month of prepaid voice-agent usage | 8,000 credits a month (~7 min), 500 chars per generation, 3 public voice slots, personal non-commercial use only; free s2.1-pro-free API model under fair use until 30 Nov 2026 | The open-source Whisper model is free to run yourself under the MIT licence; the hosted API is paid |
| Paid from | — | $5/month (Pro: 100K credits, commercial licence, instant voice cloning) | Plus $15/mo ($11/mo billed annually); API $15 per 1M UTF-8 bytes | $0.006 a minute for the whisper-1 API (removal scheduled 26 Feb 2027) |
| Has API | ✗ | ✓ | ✓ | ✓ |
| Open source | ✓ | ✗ | ✓ | ✓ |
| Free to use | ✓ | ✓ | ✓ | ✓ |
| Users | — | — | — | — |
| Founded | 2023 | 2024 | — | 2022 |
| Maker | — | — | Hanabi AI | OpenAI |
| Verdict | Kyutai is one of the most useful sources of open voice AI: Moshi for full-duplex conversation, Unmute for giving any text model a voice, and Pocket TTS for fast speech on a CPU, all with permissive code and CC-BY weights... |
Cartesia is built for real-time voice: Sonic-3.6 starts speaking in under 90 milliseconds across 44 languages, Ink-2 transcribes with built-in turn detection, and Managed Agents ties them into phone agents. Its leaderboa... |
Fish Audio is one of the strongest challengers to ElevenLabs: expressive speech with inline emotion control, voice cloning from about ten seconds of audio and support for more than 80 languages, including most Southeast... |
The open-source Whisper model is still one of the most useful pieces of speech technology available: free, MIT-licensed, multilingual and runnable offline through projects such as whisper.cpp and faster-whisper. It does... |
| Full review → | Directory listing → | Full review → | Full review → |
Comparisons cover up to 4 tools. Scores are RECATOOLS editorial assessments; verify current pricing on each vendor's site.