Skip to main content
POST
Speech-to-text

Authorizations

Authorization
string
header
required

Use a Runcrate API key with the rc_live_* prefix as the bearer token. Create one at https://www.runcrate.ai/dashboard/api-keys.

Body

multipart/form-data
file
file
required

Audio file (mp3, wav, m4a, flac, webm, mp4).

model
string
required
Example:

"openai/whisper-large-v3"

language
string

Optional hint for the spoken language as an ISO-639-1 code (e.g. 'en'; 'en-US' is read as 'en'). Unrecognised values are ignored and the language is detected, as when it is omitted.

response_format
enum<string>
default:verbose_json

Omitted: the full object (text, duration, language, segments), the same as verbose_json. json returns only {text}, as OpenAI does. text, srt and vtt return text/plain (the transcript, SubRip subtitles, WebVTT subtitles). Any other value returns 400. Billing is per minute of audio for every format.

Available options:
json,
verbose_json,
text,
srt,
vtt

Response

The transcription, shaped by response_format.

text
string
required
duration
number

verbose_json only. Audio duration in seconds.

language
string | null

verbose_json only. Detected language code.

segments
object[] | null

verbose_json only. Timed segments of the transcript.