{"slug":"transcribe-audio-url-via-mcp","question":"How can an AI agent transcribe an audio file it only has a URL for?","title_zh":"Agent 只有一个音频链接,怎么拿到文字稿?","answer":"Use a hosted speech-to-text tool: AI NetCafé's transcribe_audio MCP tool fetches the audio URL server-side and runs open-source Whisper (100 languages, self-hosted — the audio never goes to a third-party cloud). Works for voice memos, podcast clips and meeting recordings up to ~15 MB; returns plain text plus detected language.","answer_zh":"用托管的语音转文字工具:AI网吧的 transcribe_audio 会在服务端抓取音频 URL 并跑开源 Whisper(100 种语言,自托管,音频不经过第三方云)。语音备忘、播客片段、会议录音都行(≤15MB),返回纯文本和识别出的语言。","updated":"2026-08-04T19:59:47.821Z","method":"Live application tasks are run repeatedly; pass/fail, actual billed USD cost, and latency are recorded. Upstream-unavailable models are excluded from quality conclusions.","tool":{"name":"transcribe_audio","description":"Fetch an audio file from a URL and transcribe it to text with open-source Whisper (100 languages, self-hosted). Good for voice memos, podcast clips and meeting recordings up to ~15 MB.","arguments_example":{"url":"https://example.com/memo.mp3","language":"zh"},"mcp_endpoint":"https://ainetcafe.com/mcp","openapi":"https://ainetcafe.com/openapi.json"},"steps":["Get a direct, publicly fetchable audio URL (mp3/wav/m4a/ogg, ≤15 MB).","Call transcribe_audio with the url (add a language hint for faster, more accurate results).","Use the returned text; for long recordings, split the file first."],"evidence":[],"page":"https://ainetcafe.com/agent-guides/transcribe-audio-url-via-mcp","machine_readable":"https://ainetcafe.com/api/agent-guides/transcribe-audio-url-via-mcp","license":"CC BY 4.0"}