Providers
eleven_v4_turbo
The ElevenLabs text-to-speech model eleven_v4_turbo, with every setting it accepts, the allowed value and the default.
The low-latency form of ElevenLabs' v4 generation, built for live calls: it starts speaking as quickly as Flash. It speaks 85 languages and exposes stability alone.
Reached through the ElevenLabs API with OnCore's own API key, not the LiveKit Inference gateway.
Voice settings are set on the assistant, not in the voice body. Leave every one unset and ElevenLabs' own defaults apply; set any one of them and OnCore sends stability and similarityBoost too, at the same values ElevenLabs defaults to.
Settings#
Every setting is optional. Leave one out and OnCore sends nothing for it, so the Default applies. If invalid is what the provider does with a value it does not accept.
| Setting | Type | Allowed | Default | If invalid |
|---|---|---|---|---|
stability | number | 0 to 1 (0% to 100%) | 0.5 | Refused, fallback voice speaks |
Overriding on a single call#
Return this from your assistant-request webhook to change these settings for one call only. The stored voice is left untouched, and the next call starts from it again.
{
"data": {
"overrides": {
"voice": {
"settings": {
"stability": 0.5
}
}
}
}
}Send only the keys you want to change: anything you leave out keeps the value the assistant already has. Add "id" beside "settings" to swap the whole voice for another one from your catalogue at the same time.
similarityBoost, speed, style and useSpeakerBoost are not sent to this model at all. v4 Turbo is reached the same way as v3, and that route carries stability alone, so the other four are dropped before the request leaves OnCore. Nothing refuses them and nothing changes: even a speed far outside its range is accepted and has no effect.
A stability outside 0 to 1 is different: ElevenLabs refuses it, and the call carries on in OnCore's fallback voice rather than the one you chose. The caller hears a different voice and a greeting that starts a few seconds late, and nothing ends the call, so check the value if an assistant suddenly sounds like someone else.
OnCore's emotion tagging is not available on v4 Turbo, and a phonetic rule in a dictionary is applied through its plain spelling rather than its phonemes.
Next steps#
- eleven_v3: the previous ElevenLabs model.
- Cartesia: text-to-speech from Cartesia.
- Providers introduction: per-call overrides, and how a wrong value behaves across providers.