ElevenLabs releases Eleven Realtime speech-to-speech models
Original title: Elevenlabs releases two new speech-to-speech models designed for real-time AI conversations
v1 Max is billed at 98.8 percent on Big Bench Audio with a 0.57-second response; both ship on ElevenAgents.
ElevenLabs released two speech-to-speech models under the Eleven Realtime name for live conversations. The company says v1 Max scores 98.8 percent on Big Bench Audio with a 0.57-second response and that no Artificial Analysis leaderboard model is both faster and more accurate. v1 Mini is positioned for cheaper natural conversation, including detecting when a speaker has finished or is only saying “mm-hmm,” while v1 Max targets tasks such as customer support. Both are offered through ElevenAgents and an OpenAI-compatible Realtime API, priced per minute plus tokens.
Unconfirmed: The speed and accuracy ranking is ElevenLabs’ claim about the Artificial Analysis leaderboard, not an independent finding.
- v1 Max costs more than v1 Mini; exact rates are on ElevenLabs’ pricing page.
- v1 Mini is described as detecting end-of-turn and brief backchannels.
- Availability is ElevenAgents plus an OpenAI-compatible Realtime API.
- Vendors
- ElevenLabs
- Products
- Eleven RealtimeElevenAgents
From the original article
1 min read at the-decoder.com
Elevenlabs' new Eleven Realtime speech-to-speech models combine speed and accuracy. The v1 Max variant scores 98. 8 percent on the Big Bench Audio benchmark with a 0.
By Matthias Bastian · the-decoder.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.