VoiceHue is a free AI voice generator: paste your text, choose a voice, press Convert and listen or download the result as an MP3. No sign-up is needed, and one conversion takes up to 3,000 characters.
Free text to speech: paste up to 3,000 characters, pick a voice and listen in your browser. The voices load in a moment.
Create a free account and API key, then use your free allowance of 10,000 characters a month per key, in code or through MCP. No card needed.
VoiceHue runs a remote MCP server, so an AI assistant such as Claude or Cursor can turn text into speech by itself: it calls the text_to_speech tool and gets MP3 audio back. Create a free API key, add the server to your assistant with the config below, and ask it to read something aloud. The same key works with the OpenAI-compatible endpoint and with automation tools such as n8n, Make and Zapier. The free plan allows 10,000 characters a month per key.
{
"mcpServers": {
"voicehue": {
"url": "https://voicehue.com/mcp",
"headers": {"Authorization": "Bearer ${env:VOICEHUE_API_KEY}"}
}
}
}claude mcp add --transport http voicehue https://voicehue.com/mcp \ --header "Authorization: Bearer $VOICEHUE_API_KEY"
All AI tool configs: Claude Desktop, Cursor, LangChain, n8n and curl
An AI voice generator turns written text into natural-sounding speech. You do not record anything: you type or paste a script, choose a voice and get an audio file. It is used for video narration, course lessons, podcast intros, accessibility and for hearing how a text sounds before you publish it.
VoiceHue has 6 cloud voices that cover English, Chinese, Spanish, French, Japanese and Korean. Preview them on the page, filter by language and voice type, and pick the one that suits your script. A calm, steady voice suits tutorials and explainers; a brisker one suits short clips. Voices from your own device are also listed; they play in the browser but cannot be saved.
Everything on this page is free. Each text can be up to 3,000 characters (10,000 when you are signed in). Without an account you can convert 10,000 characters a day, and 30,000 characters a day when you are signed in. The audio is an MP3 you can download from the tool.
Use short sentences with one idea each, write numbers and symbols the way they should sound, and use commas and full stops where you want pauses. Listen once from start to finish; every place where the voice stumbles is a place to rewrite the sentence rather than fight the voice.
The Share button next to Download MP3 creates a link that anyone who has it can open. A link lasts 90 days, you can delete it at any time, and it has a report button. Share only text you are happy to make public, and read the licensing and copyright page on this site before you publish audio anywhere.
Yes. It is free and needs no sign-up. Without an account you can convert 10,000 characters a day, and 30,000 characters a day when you are signed in.
The 6 cloud voices cover English, Chinese, Spanish, French, Japanese and Korean. Your device may offer voices for other languages.
Yes. Choose a cloud voice, press Convert and download the MP3. Voices from your device are listen-only.
Up to 3,000 characters per conversion without an account and 10,000 when you are signed in. For a longer script, convert it in parts.
Claude and Cursor can connect to the VoiceHue MCP server at voicehue.com/mcp with a free API key and then call the text_to_speech tool. Other assistants that accept a remote MCP server or an OpenAPI description, such as ChatGPT with connectors or custom actions, can use the same endpoint and the description at voicehue.com/openapi.json, depending on what your plan of that product supports.
Create a free API key in your VoiceHue dashboard, add the MCP server voicehue.com/mcp to the agent with an Authorization: Bearer header, and let it call text_to_speech. The agent gets MP3 or WAV audio back. The developers page has copy-and-paste configs for Claude, Cursor, LangChain, n8n, Make and Zapier.
There is a free plan: 10,000 characters a month and 2,000 a day for each key, 10 requests a minute, up to 1,200 characters per call. The MCP server and the HTTP API share the same keys and limits.
Paste it into the free tool, pick a voice and press play. Cloud voices download as MP3, free within the daily limits.