Skip to main content
Rase AI Intelligence Newsletter

Creative AI Model or AI service

ElevenLabs adds separate Scribe models for recorded and live speech

Scribe v3 transcribes recorded files, while Scribe v3 Live produces text during a conversation. ElevenLabs exposes both through its Speech to Text API. [1][3]

Model access
Hosted speech-to-text access for recorded files and a separate realtime model. [1][2]
License
ElevenLabs offers the models through its hosted product and API. [2][3]

What the model does

ElevenLabs documented API access to both models on November 9, 2026, and published the public announcement on November 10. The live model can identify speaker changes during transcription. [1][3]

Why it matters

A team can choose a file model for an existing archive or a live model for captions during a meeting. The two workflows have different latency and review requirements. [1][3]

Who can use it

Application developers can use the file model for recorded interviews or the Live model for captions. ElevenLabs documents distinct API paths for the two modes. [1][3]

Key dates

  1. API access documented [1]
  2. Scribe v3 series publicly announced [3]

Ways to use it

Test the model on recordings with overlapping voices, names and numbers. For live captions, check the partial text and the final transcript, because a provisional line can change. [1][3]

Model limitations

Noise and overlapping speech can still cause transcription or speaker-label errors. The developer’s language and accuracy tests describe benchmark conditions. [2][3]

Sources and verification

Catalog evidence checked How we verify records

  1. Primary source ElevenLabs November 9 API changelog documenting Scribe v3 and v3 Live access
  2. Primary source ElevenLabs model documentation for file and live transcription
  3. Primary source Introducing Scribe v3 and v3 Live, our most accurate speech to text models

Related profiles

Creative AI

Seedance 2.5 (4K API and Flash)

ByteDance (Volcano Engine / BytePlus)

These tools generate video through developer services. The cited sources describe different variants and routes of access.

Medium confidenceEvidence checked

Creative AI

LongCat-Video-Avatar 1.6

Meituan (LongCat team)

It generates video of an avatar from supported inputs in the documented project.

High confidenceEvidence checked

Creative AI

MiniMax H3-Image

MiniMax

It generates images from prompts through the model architecture described by MiniMax.

High confidenceEvidence checked