Mistral launches Voice Agents for real-time voice applications
Mistral has released Voice Agents, a managed pipeline for two-way voice interactions in real time. Its SDK offers telephony, browser and custom-audio interfaces.
What changed
Voice Agents brings microphone capture, speech-to-text, a language model, text-to-speech and playback into a managed pipeline. The pipeline is exposed through the Voice Agents SDK. It supports telephony, browser and custom-audio interfaces.
Why it may matter for your work
This could reduce the integration work needed to build a voice-based support application. Teams may be able to use the managed pipeline rather than connecting each speech and language component themselves.
What it can actually do
Run a real-time bidirectional voice pipeline covering microphone capture, transcription, language-model processing, speech generation and playback.
Limits to check
The page links to Voice Agents pricing but states no prices or availability restrictions.
“We released **Voice Agents**, a managed real-time bidirectional voice pipeline (microphone capture, STT, LLM, TTS, playback) exposed through the Voice Agents SDK, with telephony, browser, and custom-audio interfaces.”From the linked source page