Try Eleven v4 Turbo in the Workbench
Run this model interactively, tune parameters, and compare outputs.
elevenlabs-eleven-v4-turbo
Eleven v4 Turbo is the low-latency variant of Eleven v4, producing expressive speech in more than 90 languages at about 100 ms median inference latency.
It shares Eleven v4’s audio tags, voice settings, and language coverage, so [whispering], [laughing], and similar tags direct delivery the same way. It fits voice agents, assistants, interactive characters, and bulk narration where speed and cost matter. A single request takes up to 10,000 characters.
Example request
- Minimal
- All parameters
Fetch model details
The models endpoint returns the full model object, including itsjson_request_schema.