Mistral Drops Voxtral TTS: Open-Source Voice AI in 9 Languages
French AI lab Mistral unveils an open-source text-to-speech model built on Ministral 3B, targeting enterprise voice applications.
Mistral just shipped Voxtral TTS, an open-source text-to-speech model built for enterprise deployments. It supports nine languages out of the box, including Hindi and Arabic — a notable multilingual spread for an open-source offering.
The model is based on Mistral's Ministral 3B architecture. Target use cases include voice AI assistants and enterprise workflows like customer support, where natural-sounding speech synthesis matters.
The open-source angle is the real play here. Enterprise TTS has long been dominated by proprietary systems from the usual suspects. Mistral is betting that giving companies full access to the model weights will drive adoption in a market that increasingly demands customization and data control.
Nine languages at launch is solid but not groundbreaking. The inclusion of Hindi and Arabic signals Mistral is thinking beyond Western European markets from day one.