ElevenLabs launches Eleven Realtime speech-to-speech models
ElevenLabs has launched two Eleven Realtime models, speech-to-speech systems that work as complete voice agents: v1 Mini, built for natural conversations at low cost, and v1 Max, built for complex, multi-step tasks such as customer support. Both support more than 90 languages and any of the 17,000+ voices in the ElevenLabs library. ElevenLabs says the models are compatible with OpenAI's Realtime API over WebSocket, so developers can switch by changing the endpoint and model name. On tool-use tasks, Max completed 87.1% and Mini 60.3%, so ElevenLabs recommends Max for heavy tool calling.