Audio Whisper API

Whisper Large v3

Best-in-class open speech recognition for ~100 languages, behind an OpenAI-compatible API.

Overview

Whisper large-v3 transcribes and translates speech in about 100 languages. Running in float16 with chunked batched decoding, an RTX 5090 transcribes far faster than real time.

Key features

✓
~100 languages
✓
OpenAI-compatible transcription endpoint
✓
MIT licence

Use cases

→
Meeting and call transcription
→
Podcast subtitles
→
Voice data pipelines

On this template

Runs on the Whisper API template, exposing its Transcription API at a private HTTPS address. You also get root SSH and JupyterLab.

WHISPER_MODEL openai/whisper-large-v3

Quick start

Click Deploy and pick an available GPU — only offers with enough VRAM are shown.
Wait for the status to turn running. The first start downloads the model, which can take a few minutes.
Use the connection details on your rental page:

curl

curl https://r12-abcd1234-app.gpu.clearlayer.ca/v1/audio/transcriptions \
  -H "Authorization: Bearer <app-token>" \
  -F file=@meeting.mp3 -F model=whisper-1

Python

from openai import OpenAI

client = OpenAI(base_url="https://r12-abcd1234-app.gpu.clearlayer.ca/v1", api_key="<app-token>")
with open("meeting.mp3", "rb") as f:
    print(client.audio.transcriptions.create(model="whisper-1", file=f).text)