Agent Data Tools: Speech to text and text to speech
Transcribe audio and video (SRT/VTT subtitles, an hour for $0.05) and text to speech as MP3.
我该使用它吗
质量与安全性
发现(7)
- HIGH
- MEDIUM在 transcribe 中
- MEDIUM在 transcribe_long 中
- MEDIUM在 transcribe_summary 中
- MEDIUM在 transcribe_speakers 中
- MEDIUM在 dub 中
- LOW在 voice_agent 中
基于对工具定义和协议合规性的自动分析。
上下文开销
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高,可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中:
{
"mcpServers": {
"agent-data-tools-voice": {
"url": "https://agent-data-tools.revmesh2074.workers.dev/mcp/voice"
}
}
}远程端点
https://agent-data-tools.revmesh2074.workers.dev/mcp/voicestreamable-http它能做什么
工具清单
工具(10)
🟢transcribe(url, subtitles, timestamps, language)
Transcribe speech from an audio file at a public URL, or uploaded in a POST body, using Whisper large-v3-turbo. Returns the transcript text, detected language, duration and word count; subtitles=srt or vtt adds subtitle text and timestamps=true adds timed segments and words. Accepts common formats such as MP3, WAV, M4A and OGG up to 5 MB, in 90+ languages. Use for voice notes, podcast clips, meeting snippets, subtitles and voice messages. Example: url=https://commons.wikimedia.org/wiki/Special:FilePath/En-us-hello.ogg. Price $0.01/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Absolute http(s) URL of an audio file (max 5 MB); or POST the file itself (raw body, multipart `file` part, or base64 in JSON). With no input at all this transcribes the short public sample shown in the example."
},
"subtitles": {
"type": "string",
"description": "srt or vtt: also return subtitle text"
},
"timestamps": {
"description": "true: also return timed segments and words",
"type": [
"string",
"number",
"boolean"
]
},
"language": {
"type": "string",
"description": "Spoken language code such as en, sv or de (default: detected)"
}
},
"required": [
"url"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢text_to_speech(text, lang, voice, format, sample_rate, ...)
Text to speech for a paragraph or a page: up to 2,000 characters as one MP3 in English (40 voices) or Spanish (10 voices) with Deepgram Aura-2; fr, zh, ja and ko are retired with MeloTTS (Workers AI, 20 Nov 2026). Pick a voice by name or leave the default. Use for narration, article and summary read-outs, voice replies, IVR prompts and accessibility; for a sentence of up to 300 characters ai_speech costs less. Price $0.08/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"text": {
"type": "string",
"description": "Text to speak, up to 2,000 characters"
},
"lang": {
"type": "string",
"enum": [
"en",
"es"
],
"description": "Language: en or es (Aura-2 MP3, many voices); fr, zh, ja and ko are retired (MeloTTS was removed from Workers AI, 20 Nov 2026); default en"
},
"voice": {
"type": "string",
"description": "Voice name. English: amalthea, andromeda, apollo, arcas, aries, asteria, athena, atlas, aurora, callista, cora, cordelia, delia, draco, electra, harmonia, helena, hera, hermes, hyperion, iris, janus, juno, jupiter, luna, mars, minerva, neptune, odysseus, ophelia, orion, orpheus, pandora, phoebe, pluto, saturn, thalia, theia, vesta, zeus. Spanish: sirio, nestor, carina, celeste, alvaro, diana, aquila, selena, estrella, javier. Default: luna (en), aquila (es)"
},
"format": {
"type": "string",
"enum": [
"mp3",
"wav",
"pcm",
"ulaw",
"alaw",
"opus",
"flac"
],
"description": "Audio format: mp3 (default), wav, pcm (raw 16-bit little-endian mono), ulaw or alaw (8 kHz telephony, e.g. Twilio), opus (Ogg) or flac. opus and flac only for text spoken in one part (up to 450 characters)"
},
"sample_rate": {
"description": "Sample rate in Hz for wav and pcm (8000, 16000, 24000 default, 32000, 48000) or ulaw/alaw (8000 default, 16000)",
"type": [
"string",
"number",
"boolean"
]
},
"pronunciations": {
"type": "string",
"description": "How to say words: JSON {\"Nginx\":\"engine x\",\"GIF\":\"jif\"} or Nginx=engine x, GIF=jif (up to 50, whole words, any case)"
}
},
"required": [
"text"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢transcribe_long(url, subtitles, timestamps, language)
Transcribe a podcast, meeting, interview or lecture: up to 60 minutes of MP3 or WAV (25 MB) by URL or uploaded in a POST body, 90+ languages, Whisper large-v3-turbo. Parts are joined, so subtitles=srt or vtt and timestamps=true cover the whole recording; longer audio is refused before anything runs. Use for meeting notes, podcast show notes, video subtitles and interviews; for a short clip in any format, transcribe costs less. Example: url=https://upload.wikimedia.org/wikipedia/commons/7/77/Tofino_pronunciation.mp3. Price $0.05/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Absolute http(s) URL of an MP3 or WAV file (max 25 MB, 60 minutes); or POST the file itself (raw body, multipart `file` part, or base64 in JSON)"
},
"subtitles": {
"type": "string",
"description": "srt or vtt: also return subtitle text for the whole recording"
},
"timestamps": {
"description": "true: also return timed segments and words",
"type": [
"string",
"number",
"boolean"
]
},
"language": {
"type": "string",
"description": "Spoken language code such as en, sv or de (default: detected per part)"
}
},
"required": [
"url"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢transcribe_summary(url, language, transcript)
Summarize a voice note, meeting, call or podcast in one call: transcribes up to 60 minutes (MP3/WAV up to 25 MB; M4A, OGG or WebM up to 5 MB) by URL or upload, then returns a title, summary, key points, action items with owner and due date, open questions and the transcript, in the recording's language. Use for meeting notes, voicemail to CRM and call notes. Example: url=https://upload.wikimedia.org/wikipedia/commons/7/77/Tofino_pronunciation.mp3. Price $0.05/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Absolute http(s) URL of the recording (MP3/WAV up to 25 MB and 60 minutes; M4A, OGG or WebM up to 5 MB); or POST the file itself (raw body, multipart `file` part, or base64 in JSON)"
},
"language": {
"type": "string",
"description": "Spoken language code such as en, sv or de (default: detected)"
},
"transcript": {
"description": "false: leave the full transcript out of the answer (default true)",
"type": [
"string",
"number",
"boolean"
]
}
},
"required": [
"url"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢transcribe_speakers(url, language, redact, subtitles)
Transcribe a meeting, interview, call or podcast with speaker labels: each turn says which speaker talked, with start and end times (Deepgram Nova-3). MP3/WAV up to 15 minutes, other formats up to 2 MB, by URL or upload. Optional redaction of personal data (pii), card numbers (pci) or numbers, and subtitles with the speaker in each line. Use for meeting minutes, interview transcripts, call reviews and podcast show notes. Example: url=https://upload.wikimedia.org/wikipedia/commons/7/77/Tofino_pronunciation.mp3. Price $0.10/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Absolute http(s) URL of the recording (MP3/WAV up to 15 minutes; M4A, OGG, WebM or FLAC up to 2 MB); or POST the file itself (raw body, multipart `file` part, or base64 in JSON)"
},
"language": {
"type": "string",
"description": "Spoken language code such as en, es, de or fr (default: detected)"
},
"redact": {
"type": "string",
"description": "pii (names, addresses, phone numbers and other personal data), pci (card numbers) or numbers: replaced in the transcript"
},
"subtitles": {
"type": "string",
"description": "srt or vtt: also return subtitles, each line starting with its speaker"
}
},
"required": [
"url"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢subtitles_translate(subtitles, to, from, url)
Translate an SRT or WebVTT subtitle file into another language cue by cue, keeping every timing and the file format, so the result drops straight into a video player or editor. Up to 400 cues (40,000 characters) as text or by URL; the source language is detected unless given. Meta M2M100 translation, 100 languages. Use to subtitle videos, courses and podcasts in more languages; pair with transcribe_long, which writes the first SRT. Price $0.02/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"subtitles": {
"type": "string",
"description": "The SRT or VTT text (or give url)"
},
"to": {
"type": "string",
"description": "Target language code such as sv, de, es, fr or ja"
},
"from": {
"type": "string",
"description": "Source language code (default: detected from the text)"
},
"url": {
"type": "string",
"description": "URL of an .srt or .vtt file, instead of subtitles"
}
},
"required": [
"subtitles",
"to"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢dub(url, to, from, voice)
Dub a clip: speech in any of 90+ languages is transcribed, translated and spoken again, with translated SRT on the original timings. New voice in English or Spanish (Aura-2 MP3, pick a voice, up to 2,500 characters); fr, zh, ja and ko are retired with MeloTTS (Workers AI, 20 Nov 2026). MP3/WAV up to 10 minutes or other formats up to 2 MB, by URL or upload. Keeps the order, not exact timing. Use for course clips and product videos. Example: url=https://upload.wikimedia.org/wikipedia/commons/7/77/Tofino_pronunciation.mp3, to=es. Price $0.10/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Absolute http(s) URL of the recording (MP3/WAV up to 10 minutes; M4A, OGG or WebM up to 2 MB); or POST the file itself (raw body, multipart `file` part, or base64 in JSON)"
},
"to": {
"type": "string",
"enum": [
"en",
"es"
],
"description": "Language of the new voice: en or es (Aura-2; fr, zh, ja and ko are retired with MeloTTS, 20 Nov 2026)"
},
"from": {
"type": "string",
"description": "Spoken language of the recording (default: detected)"
},
"voice": {
"type": "string",
"description": "Aura-2 voice for the new track (default luna for en, aquila for es)"
}
},
"required": [
"url",
"to"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢voice_agent(text, url, instructions, history, lang, ...)
One turn of a voice assistant: send the user's voice note (or text), your instructions and the earlier turns; get back what was heard, a short reply, the reply spoken in English or Spanish and the updated history to send next time. Speech up to 2 minutes (MP3/WAV, or 1 MB of other formats); replies up to 600 characters. Use for voice bots in Telegram or WhatsApp, phone menus, kiosks and hands-free helpers. Example: text=What time do you open on Saturday?. Price $0.05/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"text": {
"type": "string",
"description": "The user's turn as text (or send audio with url or a POSTed file)"
},
"url": {
"type": "string",
"description": "Absolute http(s) URL of the user's spoken turn (MP3/WAV up to 2 minutes, other formats up to 1 MB); or POST the file; or send text instead"
},
"instructions": {
"type": "string",
"description": "Who the assistant is and what it knows (up to 2,000 characters)"
},
"history": {
"type": "string",
"description": "Earlier turns as a JSON array of {role, content}, as returned in history by the last call"
},
"lang": {
"type": "string",
"enum": [
"en",
"es"
],
"description": "Reply language and voice: en or es (Aura-2 MP3); fr, zh, ja and ko are retired, so a request for them is a 422 (MeloTTS was removed from Workers AI, 20 Nov 2026); default en"
},
"voice": {
"type": "string",
"description": "Aura-2 voice for an en or es reply (default luna for en, aquila for es)"
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢text_to_speech_captions(text, lang, voice, captions)
Text to speech that comes with subtitles: up to 2,000 characters spoken as one MP3 in English or Spanish (Deepgram Aura-2, pick a voice), plus SRT or VTT captions whose timings come from the generated audio itself, sentence by sentence. Ready for a video editor, no transcription step. Use for narrated videos, social clips, e-learning and accessible audio with captions. Price $0.08/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"text": {
"type": "string",
"description": "Text to speak, up to 2,000 characters"
},
"lang": {
"type": "string",
"enum": [
"en",
"es"
],
"description": "Language: en or es (default en)"
},
"voice": {
"type": "string",
"description": "Aura-2 voice name (default luna for en, aquila for es)"
},
"captions": {
"type": "string",
"description": "srt (default) or vtt"
}
},
"required": [
"text"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢dialogue(script, lang, voices, captions)
Turn a script into a multi-voice conversation: write one line per turn as "Name: text", each speaker gets their own natural voice (Deepgram Aura-2, English or Spanish, or pick voices), and the whole dialogue comes back as one MP3 with SRT captions naming the speaker, timed from the real audio. Up to 6 speakers and 2,500 characters. Use for podcast intros, language-lesson dialogues, explainer videos, ads and game or story scenes. Example: script=Anna: Welcome back to the show! Ben: Thanks, it is great to be here.. Price $0.10/call via x402 USDC or a card key (/credits); failed calls free. Reads or computes only and changes nothing; each call is charged, so repeats are not free.
输入模式
{
"type": "object",
"properties": {
"script": {
"type": "string",
"description": "One turn per line as Name: text (up to 6 speakers, 2,500 characters)"
},
"lang": {
"type": "string",
"enum": [
"en",
"es"
],
"description": "Language: en or es (default en)"
},
"voices": {
"type": "string",
"description": "Voice per speaker, as JSON {\"Anna\":\"luna\"} or Anna=luna, Ben=orion (default: a different voice each, in order)"
},
"captions": {
"type": "string",
"description": "srt (default) or vtt"
}
},
"required": [
"script"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}社区
证据