Skip to main content
POST
Create realtime transcription session.

Authorizations

Authorization
string
header
required

Bearer JWT token or API Key (sk-*).

Body

application/json
language
string

Optional ISO-639-1 language hint for the spoken audio, for example "en".

Maximum string length: 8
Example:

"en"

prompt
string

Optional extra transcription prompt or vocabulary. The backend always adds the assistant coding/business vocabulary baseline.

Maximum string length: 1000
expiresAfterSeconds
number
default:120

Client secret lifetime in seconds. Keep this short because the token is usable from the browser.

Required range: 10 <= x <= 600

Response

Realtime transcription session created.

clientSecret
string
required

Short-lived credential for browser audio transcription.

expiresAt
number
required

Unix timestamp in seconds when the client secret expires.

model
string
required

Identifier of the transcription model selected by the instance.

sessionType
string
required

Audio transcription session type.

Example:

"transcription"

Last modified on September 5, 2026