# Whisper Transcriber — audio transcription & speech to text API An ai tool your agent can run on one prepaid apifare key. Page: https://apifare.com/tools/amanatools/whisper-transcriber ## What you need An apifare key with credits (sign up: POST https://apifare.com/v1/signup {"email": "..."}; new accounts get 500 free credits). No other account or key. ## Inputs - `media_urls` (array, required): Media URLs - `model` (string): Model (e.g. "tiny") - `language` (string): Language (e.g. "auto") - `task` (string): Task (e.g. "transcribe") - `rss_episodes` (integer): Episodes per RSS feed (e.g. 1) - `max_minutes_per_file` (integer): Max minutes per file (e.g. 120) The most relevant fields; the search result's input_schema lists all of them. ## Price 5 credits reserved per run, settled to the run's true cost, never more than 20 credits (1 credit = $0.01). Failed runs are refunded. Exact numbers: https://apifare.com/capabilities ## Run it 1. Find it: MCP search_apis, or GET https://apifare.com/v1/search?q=Whisper%20Transcriber%20%E2%80%94%20audio%20transcription%20%26%20speech%20to%20text%20API with Authorization: Bearer . Take the result named "Whisper Transcriber — audio transcription & speech to text API"; it carries a proxy_route. 2. Run it: MCP proxy_call with that proxy_route and the input JSON (apifare-mcp over stdio), or POST the proxy_route over HTTP with the input JSON as the body and the same Authorization header. 3. On HTTP 402, relay the body's human_readable field to your human verbatim, wait, then retry the same call. Every other error and its fix: https://apifare.com/errors.md