Product documentation

Speech Transcription API

Transcribe authorized audio into clean text with usage metered by duration. This focused audio engine template defines one documented request contract, an enforceable package, and locked technical settings. It remains unavailable for purchase until UrTheProduct verifies an exact provider mapping, rights, reliability, cost ceiling, and final price.

This page is the public integration contract. Purchase is enabled only when the exact provider, cost ceiling, final price, and Whop plan are approved; internal prompts and provider details are never part of the public request.

Operation and route

audio.transcribe
POST /v1/audio/transcribe

Package configuration

Standard package

20 audio hours · strict file and format limits

20 audio hours
Expires 30 days after fulfillment
5 requests per minute; 1 concurrent request(s)
Status: launch configuration review

Locked configuration

{
  "mime_types": [
    "audio/mpeg",
    "audio/wav",
    "audio/mp4"
  ],
  "timestamps": false,
  "max_total_hours": 20
}

Request example

{
  "language": "en",
  "input_upload_id": "upl_required"
}

Restrictions

Use only with content you own or are authorized to process. Do not use the engine for deception, abuse, infringement, unlawful surveillance, or attempts to bypass its locked settings.

Implementation path

Store the revealed key in a server environment variable.
Use this package's exact route and strict request body.
Send bearer authentication and one idempotency key per logical POST.
Handle quota, validation, rate, provider, and retry cases from the shared references.

Open the complete quickstart →

Review errors, quota effects, and safe retries →