Available workloads

Transcription

Use whisper_ane for speech-to-text. It returns timestamped segments and detects the spoken language automatically. Pass a URL to the audio file as input_uri.

python

job = cc.submit(
    "whisper_ane",
    {"input_uri": audio_url},
    model_id="whisper-large-v3-turbo",
    data_class="public",
    marketplace_execution_risk_acknowledged=True,
)