XTTS-v2
An open-weights multilingual TTS and voice cloning model covering 17 languages at 24kHz, still widely used for local voice cloning.
- Resolution
- 24kHz
- Weights
- Downloadable
- API
- No
- Commercial use
- The licence forbids commercial use
Coqui Inc. shut down in January 2024, so the model and the coqui-ai/TTS toolkit are community-maintained. It is released under the Coqui Public Model License, not a standard open-source licence.
Coqui Public Model License, which is not OSI-approved and is reported by multiple independent analyses to restrict use to non-commercial purposes. Coqui has shut down, so no licensor remains to grant a commercial exception.
Specification
- What it takes in
- text, audio
- What it makes
- audio
- Maximum resolution
- 24kHz
- Controls
- voice-cloning, speed-control
- Weights
- Downloadable
- API
- No
- Pricing
- Free to download and self-host; compute cost only. Coqui Studio and its API were wound down at the turn of 2024, so there is no hosted first-party option.
- Released or updated
- Coqui Inc. shut down 2024-01
- Languages
- 17, not individually listed by the vendor
Strengths
- 17 languages with zero-shot cloning, fully local
- Mature, heavily documented, large community around coqui-ai/TTS
- 24kHz output
- No per-character cost
Weaknesses
- CPML is non-commercial - a hard blocker for client or commercial video work
- No path to buy a commercial licence because Coqui Inc. is dissolved
- Original licence URL is dead, so the exact terms are hard to verify
- Unmaintained by the original vendor since January 2024
- Outclassed on quality by newer MIT and Apache models such as Chatterbox and Kokoro
Best for
- research and personal experimentation
- internal prototyping only
- nothing commercial
Where the sources disagree
Recorded rather than resolved. Picking the tidier number would hide the fact that the vendor's own pages do not agree.
- The Hugging Face model card names CPML but does not state whether commercial use is permitted. Two third-party legal summaries state CPML is non-commercial. The canonical licence text (coqui.ai/cpml) is offline. commercialUseAllowed is set to false on the conservative reading; if this matters to a production, get independent legal advice rather than relying on this entry.
- Shutdown date is reported as 3 January 2024 by third parties; the GitHub discussion thread confirms the shutdown but the exact date was not read from it.
What to understand before using it
Explainers naming the mechanisms this model's own documentation mentions.
See it from the vendor
Shotcraft hosts no sample output. Judging a model means watching its own showreel with your own eyes, and a clip re-hosted here would be a clip we chose.
Nearby in the directory
Azure AI Speech text-to-speech (Neural, Neural HD, MAI voices)
Microsoft's enterprise speech synthesis service, with 200+ prebuilt neural voices across 80+ locales plus tiers for HD voices, multilingual voices, mu
Voice and speech
Sonic 3.5
Cartesia's current TTS model, documented at sub-90ms first-byte latency across 42 languages. It is tuned for conversational delivery and for reading a
Voice and speech
Deepdub (eTTS, Speech-to-Speech, Live)
Enterprise AI voice and dubbing stack covering expressive text-to-speech with emotional control, speech-to-speech voice conversion and translation, vo
Voice and speech