#how can i control speed for v3 via the api? As far as i see there is no speed parameter for v3 and

1 messages · Page 1 of 1 (latest)

icy robinBOT
#
AI-generated Answer for robertschonholz

Short answer: set voice_settings.speed when you call Text-to-Speech with model_id=eleven_v3 (range 0.7–1.2; default 1.0). Example body fields:

  • model_id: "eleven_v3"
  • voice_settings: { "speed": 0.85 }

This overrides the voice’s stored speed just for that request. (elevenlabs.io)

If you want to persist speed for a voice, update the voice’s settings:
POST /v1/voices/:voice_id/settings/edit with speed set to your value. (elevenlabs.io)

Notes

  • Speed control is available for all models and voices and supports 0.7–1.2 (values <1 slow down; >1 speed up). (help.elevenlabs.io)
  • v3 also allows directional “audio tags” like [slowly], but their effect is voice- and context-dependent and can be reduced by high “Robust” stability; use the speed parameter for deterministic pacing. (elevenlabs.io)

If you’d like, I can paste a minimal cURL or SDK example with voice_settings.speed for v3.

Sources:
Create speech | ElevenLabs Documentation
Edit voice settings | ElevenLabs Documentation
Can I change the pace of the voice? – ElevenLabs
Prompting Eleven v3 (alpha) | ElevenLabs Documentation

rain condor
#

i tried but it doesnt work with v3 model

icy robinBOT
#
AI-generated Answer for robertschonholz

Short answer: Eleven v3 (Alpha) does support speed control, but its pacing is primarily driven by inline audio tags. If the numeric speed setting seems to have little effect, use tags to direct pacing.

What works today

  • API speed parameter: Send voice_settings.speed in your Create speech request (range 0.7–1.2; 1.0 = normal). This overrides the voice’s stored settings for that request. Example body: { "model_id": "eleven_v3", "text": "...", "voice_settings": { "speed": 0.85 } }. (elevenlabs.io)
  • v3-specific pacing via tags: Put tags directly in the text, e.g., “[slowly] …”, “[rushed] …”, “[deliberate] …”. v3’s prompting guide states that speed is controlled through audio tags, and the v3 page shows “[slowly]” in examples. (elevenlabs.io)

Why your v3 call may look unchanged

  • v3 prioritizes delivery directives from tags and the chosen voice; numeric speed can have a subtler effect than on earlier models. The v3 prompting guide notes tag effectiveness depends on the voice and settings. Try adding tags and avoid extremely “Robust” stability if you want stronger responsiveness to directives. (elevenlabs.io)

Reference ranges and availability

  • Speed is available across Text to Speech (web and API) with valid values 0.7–1.2. (help.elevenlabs.io)

If you want, paste your JSON body and I’ll check that model_id is eleven_v3 and voice_settings.speed is being sent correctly. (help.elevenlabs.io)

Sources:
Create speech | ElevenLabs Documentation
Prompting Eleven v3 (alpha) | ElevenLabs Documentation
Prompting Eleven v3 (alpha) | ElevenLabs Documentation
Can I change the pace of the voice? – ElevenLabs
What is Eleven v3 (Alpha)? – ElevenLabs

rain condor
#

is there a page for v3 api docs?