Skip to main content
POST
Create a speech turn

Authorizations

Authorization
string
header
required

Workspace API key. Create one via the dashboard or POST /v1/auth/keys.

Body

application/json
model
string
required

Exact realtime slug, or s2s/auto for managed feature routing.

input_audio
object
required

Base64 audio (JSON requests only). Mutually exclusive with the multipart file field.

output_audio
object
voice
string
default:auto

Managed routes require auto — native voice ids are not portable.

feature
enum<string>

Case-sensitive capability to prioritize. Valid ONLY with model s2s/auto; rejected for pinned models.

Available options:
Emotion Understanding,
Emotion Alignment,
Expressive Robustness,
Voice naturalness,
Problem redirecting
instructions
string

System guidance for how the model should answer the turn.

temperature
number
max_output_tokens
integer
metadata
object
provider
object

Per-request routing constraints, applied before ranking.

options
object

Response

OK.

id
string
object
string
Example:

"audio.speech_to_speech"

created
integer
model
string
resolved_model
string
provider
string
audio
object
transcript
string

Transcript of the model's spoken answer.

input_transcript
string | null

Transcript of your audio; null when the model does not return one.

usage
object

Normalized usage. cost is in USD; Naive credits are reported separately as credits_used.

credits_used
number
credits_remaining
number