Open-weight text-to-speech lab behind Miso TTS, an 8B emotive conversational voice model.
Miso Labs builds Miso TTS 8B, an open-weight text-to-speech model for high-quality conversational speech. The model uses a Sesame-style architecture with a large backbone and a smaller audio decoder, cloning voices from short prompt audio for English speech. It ships under a modified MIT license with watermarking by default and a no-impersonation condition. The code and weights are published on GitHub and Hugging Face under the MisoLabsAI and MisoLabs accounts.
Text-to-Speech
Open-source conversational TTS models
HQ COUNTRY
United States
FOUNDED
2025
DEPLOYMENT
API, On-premise
PRICING MODEL
Open source
COMPANY STAGE
Seed
Is this your company? To claim this listing or update its details, contact us at sunil@uniocommunity.com
