← All genres
VOICE
Voice & speech
Speech synthesis, cloning and dubbing. Latency matters as much as realism.
5 models tracked · 8 releases on record · last movement 3w from Voxtral
Where the genre stands
Ordered by most recently shipped
ModelVersion
- VoxtralMistral AIEUMistral's speech family: transcription, text-to-speech and a voice-agent SDK.3
- Fish Audio DramaFish AudioCNOpen Weights35B MoE emotive speech model with 16-character multi-speaker synthesis.3.0
- Sonic 2CartesiaUSUltra-low latency streaming voice engine with 85ms TTFB for conversational agents.2.0
- MoshiKyutaiEUOpen WeightsFirst real-time native speech-to-speech multimodal dialogue model with 160ms latency.1.5
- ElevenLabs TTSElevenLabsGlobalThe default for expressive speech synthesis and dubbing.v4
Everything that shipped
- 3wVOICEVoxtralMistral AIEUNEW DROPVoxtral 3 brings new speech-to-text and text-to-speech models and a voice-agent SDK.3
- 3wVOICEFish Audio DramaFish AudioCNNEW DROPFish Audio drops Drama 3: 35B MoE emotive speech model with 16-character multi-speaker synthesis.3.0
- 1moVOICESonic 2CartesiaUSNEW DROPCartesia Sonic 2 sets real-time voice benchmark with sub-90ms full-duplex conversational synthesis.2.0
- 1moVOICEMoshiKyutaiEUMoshi 1.5 drops latency to 110ms with multilingual conversational turn-taking and emotional mirroring.1.5
- 1moVOICEFish Audio DramaFish AudioCNFish Speech 2.0 introduces sub-100ms streaming TTS.2.0
- 1moVOICEElevenLabs TTSElevenLabsGlobalEleven v4, ElevenLabs’ most emotive model, launches with a new architecture.v4
- 1yVOICEElevenLabs TTSElevenLabsGlobalv3 adds audio tags for directed emotional delivery.v3
- 2yVOICEMoshiKyutaiEUKyutai open-sources Moshi, achieving conversational speech-to-speech without STT/TTS cascade.1.0