← Back to Voice models
ElevenLabs
Eleven Realtime v1 Mini
Released Oct 19, 2026
ElevenLabs' default, fastest and cheapest Eleven Realtime model, for assistants, tutors and interactive characters.
- Price per minute
- $0.080/min
- Inputs
- audio, text
- Tool calling
- Yes
- Languages
- 90
- Voices
- —
- Big Bench Audio
- 75.9%
- Access
- Generally available
- Licence
- Proprietary
Strengths & weaknesses
Strengths
- Full-Duplex-Bench 95.8%, higher than Max
- Input tokens $0.20 per 1M
- OpenAI-Realtime-compatible API
Weaknesses
- Big Bench Audio 75.9% (AA), far below Max
- Tool-use success 60.3%, vs 87.1% for Max
Not notable: #18 on AA S2S Index (64.4) and #26 on Big Bench Audio (75.9%). The text input modality is inferred from the OpenAI-compatible Realtime API; the blog only says audio in, audio out.
Links
API model strings
- elevenlabs
eleven_realtime_v1_mini