Skip to main content
GET
List Speech-to-Speech Voices
Read only Returns Speech to Speech voices with the languages and realtime LLM paths each voice supports. Use this endpoint when building a Speech to Speech voice selector or validating an agent configuration before saving it.

When to use this

  • Building S2S agents: agents with pipeline_type: "Speech_To_Speech" can only use voices from this list, not the general List Voices catalog.
  • Voice selectors: populate the voice dropdown in your own tooling when the operator has chosen the Speech to Speech pipeline.
  • Pre-save validation: confirm the voice, language, and realtime LLM combination is supported before creating or publishing the agent.
The API-key response is a root JSON array containing the full voice list without pagination metadata. Each voice carries id, name, provider, provider_voice_id, gender, accent, icon, recording, languages (each with a lang_ id, name and code) and llms (each with an llm_ id, name, provider and provider_id).

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Response

Speech to speech voices returned successfully.

id
string
required
Example:

"voice_mj18r4tq2w9xkp"

name
string
required
Example:

"Ash"

provider
string | null
required
Example:

"openai"

provider_voice_id
string | null
required
Example:

"ash"

gender
string
required
Example:

"male"

accent
string | null
required
Example:

"american"

icon
string | null
required
recording
string | null
required
languages
object[]
required
llms
object[]
required