Skip to content

ModelsOpenAI

GPT Transcribe

Facts

Model id
openai/gpt-transcribe
Modality
Audio · Speech to text
Endpoints
/v1/audio/transcriptions
Context window
8K
Maximum output
Not published
Data policy
No training · No zero retention
List price
0.297 USD per hour of audio

Call it

Send the id as model. The quickstart has the key and the SDK setup; models, limits and data policy explains each field above.

bash
curl https://api.sylphx.com/v1/audio/transcriptions \
  -H "Authorization: Bearer $SYLPHX_API_KEY" \
  -F model="openai/gpt-transcribe" \
  -F file="@meeting.m4a"