AnyToSpeech MCP - Connect ChatGPT, Claude, Cursor & More to Your Account
Model Context Protocol

Your AI assistant.
Full AnyToSpeech power.

Connect ChatGPT, Claude, Cursor, Perplexity or Meta Muse to your AnyToSpeech account. Convert text, transcribe meetings, clone voices and generate podcasts, without copy-pasting or a separate API bill.

ChatGPT Claude Meta Muse Perplexity Cursor VS Code Any MCP client
Get started free See how it works

Uses your existing plan, credits & transcription minutes · Sign in, no key to copy · Disconnect anytime

28
MCP tools
Sign in
No key to copy
Your plan
Same billing as the website
Scoped
Disconnect anytime
What you get

Everything AnyToSpeech does — inside your AI chat

Not a stripped-down API. The same production features your account already has, exposed as tools your assistant can call.

Text & documents → natural speech

Generate audio from plain text, PDFs, DOCX, images with text, SRT subtitles, or multi-speaker scripts. Perfect for narration, accessibility, and content repurposing.

speech_synthesize speech_from_document speech_from_image speech_multi_speaker
Text to speech in AnyToSpeech

Transcribe & translate anything

Submit audio or video for transcription, summarize recordings, translate text, audio, images, or media URLs — all metered against your transcription minutes.

transcription_create transcription_summarize translation_audio translation_image
Transcription and translation

Clone, design & transform voices

Create custom voices from samples, design new voices with AI, run speech-to-speech conversion, and manage your voice library — all from a chat prompt.

voice_clone voice_design_publish speech_to_speech
Voice cloning

Podcasts on demand

Generate full podcast episodes from a topic, upload a file, or plan a series. Your assistant handles scripting and audio — you keep publishing.

podcast_generate podcast_from_file podcast_series_plan
Podcast generation
Full toolkit

Six capability areas, one connector

Approve only the permissions each assistant needs, and see every connected app in one place.

Speech generation

Text, documents, images, SRT, and multi-speaker dialogue → downloadable audio.

Transcription

Upload or link media, poll job status, get transcripts and AI summaries.

Translation

Translate text, audio, images, and transcripts across languages.

Voice studio

Clone, design, test, convert, and delete custom voices you own.

Podcasts

Single episodes, file uploads, and multi-episode series planning.

Account & jobs

Check credits and minutes, list voices, track async jobs, cancel runs.

Setup

Connected in under two minutes

No developer portal. No separate API subscription. Just your AnyToSpeech account.

1

Create a free account

Sign up for AnyToSpeech if you haven't already. Free tier included — upgrade anytime for more credits and minutes.

2

Paste the server URL

Add the AnyToSpeech server URL in ChatGPT, Claude, Cursor or VS Code, sign in and click Allow. No key needed. Perplexity and Meta Muse use an integration key you create in one click.

3

Ask your assistant

“Transcribe this file.” “Read this PDF aloud.” “Make a podcast about…” Usage bills to your plan automatically.

Set up AnyToSpeech in your assistant

ChatGPTAdd AnyToSpeech as a custom app in developer mode. About 3 minutes. ClaudeAdd a custom connector. Works on every Claude plan, including Free. Meta MuseConnect through a Muse custom connector while our directory listing is in review. PerplexityAdd AnyToSpeech as a remote connector with an integration key. CursorOne line in mcp.json, then sign in. Narrate docs and changelogs from your editor.

ChatGPT guides

Read a PDF out loudGet the whole document as audio, not a 60-second clip. Text to speech as MP3Turn any ChatGPT answer or pasted text into a downloadable file. Create a podcastTwo hosts, natural back-and-forth, from a topic or a file. Clone your voiceMake ChatGPT's output sound like you.

Claude guides

Read a PDF out loudListen to long reports and papers Claude has already read. Add text to speechGive Claude a voice: 200+ voices, any language, MP3 output. Create a podcastHave Claude research the topic and AnyToSpeech voice it. Clone your voiceUse your own voice for everything Claude writes.

Use AnyToSpeech Inside Your AI Assistant →

Create account & connect
Trust

Built for your account, not ours

Unlike generic API keys, MCP connects to your subscription, with controls you can revoke anytime.

Disconnect anytimeEvery signed-in app and integration key is listed on your integrations page. Disconnect or revoke and access stops immediately.
Scoped permissionsChoose exactly what each assistant can do: text to speech only, transcription only, and so on.
Your plan & creditsNo surprise API invoice. Same limits and billing as the AnyToSpeech website.
Token-only API accessMCP endpoints require Bearer tokens — browser sessions cannot auth the API.
FAQ

Common questions

What is AnyToSpeech MCP?

It's a Model Context Protocol server that lets AI assistants (ChatGPT, Claude, Cursor, VS Code, Perplexity, Meta Muse and more) call AnyToSpeech on your behalf: text to speech, transcription, translation, voice cloning, podcasts and more.

Do I need a separate API subscription?

No. MCP usage draws from your existing AnyToSpeech account — the same credits, transcription minutes, and plan limits you use on the website.

How do I connect Cursor?

Click Add to Cursor on the integrations page, then click Connect next to AnyToSpeech in Cursor Settings → Tools & MCP and sign in. No key needed.

Does it work with ChatGPT, Claude, Perplexity and Muse?

Yes. In ChatGPT and Claude, add a custom connector with the server URL, sign in to AnyToSpeech and click Allow. No key needed. Perplexity and Meta Muse use an integration key, which you create in one click on the integrations page.

Is my data safe?

You approve exactly which permissions each app gets. Signed-in apps appear under Connected apps and keys under Integration keys on the integrations page. Disconnect or revoke anytime and access stops immediately.

Where is the MCP server URL?

https://anytospeech.com/api/mcp — paste it into your assistant and sign in. It's also on your integrations page.

Stop copy-pasting into AI.
Start creating with voice.

Connect AnyToSpeech once. Your assistant handles speech, transcription, and podcasts from chat.

Get started — it's free