Speech to text

Upload an audio file and get a transcript back.

POST/v1/audio/transcriptions

Send a multipart form upload. Audio models cannot be used on the chat endpoint.

FORM FIELDS

Parameter	Type	Description
`file`required	`file`	The audio to transcribe. Maximum 25 MB; larger files return 413.
`model`required	`string`	An audio model ID beginning with `sovereign/` that is enabled for your organization. Unknown models return 400; models you may not use return 403.
`response_format`	`string`	`json` (default), `text` or `verbose_json`. Anything else returns 422.

curl https://api.eggsplain.com/v1/audio/transcriptions \
  -H "Authorization: Bearer $EGGSPLAIN_API_KEY" \
  -F model="sovereign/<audio-model>" \
  -F response_format=json \
  -F file=@meeting.wav

Notes

The upload is size-checked while it is read, so a missing or wrong Content-Length does not bypass the limit.

Ask your organization admin which audio models are enabled.