media_transcribe
Speech to text for AI agents, $0.30 per file: MP3, WAV, M4A, OGG, FLAC or WebM audio (max 5 MB, about 5 minutes) to a transcript with timed segments, detected language and optional WebVTT subtitles. One call: send a url or base64 — no upload step, no polling. No speech found means a 400 and you pay nothing. USDC on Base, no account, no API key.
Tool call template
This tool declares no required parameters. Add optional parameters from the input schema as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {},
"name": "media_transcribe"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| url | No | http(s) link to the file; we fetch it. Send this OR file_base64. | |
| filename | No | Optional original filename of file_base64 (its extension helps detect the type) | |
| language | No | Optional 2-letter language code (e.g. 'en'); omit for auto-detect | |
| media_type | No | Optional MIME type of file_base64 (helps tell DOCX/XLSX apart) | |
| file_base64 | No | The file bytes, base64-encoded. Send this OR url. | |
| include_vtt | No | Also return WebVTT subtitles |