ElevenLabsAudio modelActive
Scribe v3
Summary
ElevenLabs released Scribe v3 for recorded speech on November 10, 2026. API model ID `scribe_v3` uses the Speech to Text endpoint for batch transcription. The model supports 90+ languages, diarization for up to 32 speakers, word- and character-level timestamps, and up to 1,000 keyterms.
ElevenLabs offers the model through its API and web demo. The pricing page lists USD 0.20 per hour of processed audio. Public downloadable weights are not documented.
Similar models
More models of the same kind as Scribe v3, chosen by model line, developer, size, price and release date.
- Scribe v3 LiveElevenLabs · same developer
- Eleven Music v2.5ElevenLabs · same developer
- SFX v3ElevenLabs · same developer
- Eleven Realtime v1 MaxElevenLabs · same developer
- Eleven Realtime v1 MiniElevenLabs · same developer
- Eleven v4ElevenLabs · same developer
- Eleven v4 TurboElevenLabs · same developer
- Scribe v2 RealtimeElevenLabs · same developer
More about Scribe v3
All specifications
Architecture
Input and output
- Capabilities
- Supports 90+ languages, diarization for up to 32 speakers, word- and character-level timestamps, audio-event tags, and up to 1,000 keyterms. Number formatting and expected secondary languages can be specified.
- Source: Capabilities
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
- Access
- Available through ElevenAPI. The documentation also provides a web demo for Speech to Text.
- Source: Access
- Shutdown
- None announced
- no source
API
Further details
- Weights and license
- Hosted proprietary service; public downloadable weights are not documented.
- Source: Weights and license
Prices
- API price
- USD 0.20 per hour of processed audio.Source