# AI voice cloning: clone your voice online and read any script

Record 10 to 30 seconds of speech or upload an audio or video clip, save your custom AI voice clone, and turn any script into natural MP3 speech across 30+ languages with Audo's free online voice cloner.

## How it works

1. Record 10 to 30 seconds on your microphone, or upload a WAV, MP3, M4A, FLAC, OGG, MP4, or MOV clip (5 seconds to 3 minutes), and confirm you have permission to use the voice.
2. Keep Remove background noise on to strip steady room hiss without changing your vocal timbre, and optionally name and save the voice to your account for one-click reuse.
3. Type or paste your script (with optional delivery tags like [whispers] or [small laugh]), generate the speech, and download the 192 kbps MP3.

## What you can make with Audo's AI voice cloner

One voice sample powers instant text-to-speech readings, podcast line fixes, multilingual dubs, and automated API voiceovers:

### Clone my voice for text to speech: Fix podcast pickups, YouTube narration, and course lessons without re-recording

When a product name changes or a sentence needs a clean retake, type the updated line into Audo's voice cloning text to speech tool instead of setting up your microphone again. Save your voice once so every future script reads in the same tone, or pair it with preset studio voices in Text to speech.

### Clone voice from audio or video: Clone a voice directly from an MP3, WAV, iPhone voice memo, or MP4 video

You do not need a studio session to clone a voice from an audio file or video clip. Drop in a 10 to 30 second single-speaker recording—or isolate a clean speaking clip first with Voice isolator or Remove music from video—and Audo extracts the vocal profile automatically.

### Multilingual voice cloning: Speak 30+ languages in your own voice from a single recording

Record your sample once in English (or any supported language) and paste translated scripts in Spanish, French, German, Japanese, Korean, Mandarin, Hindi, Arabic, or Portuguese. Combine it with Audio translator to translate a transcript and read it back in your own voice.

### Voice cloning API & MCP: Automate custom voiceovers from code, Claude, ChatGPT, and Cursor

Pass a reference file or a saved voice_id to clone_voice over Audo's REST API or MCP server. Coding agents and workflows can list saved voices with list_voices and generate narrated changelogs, demos, and audio lessons on demand—see Developers.

## Instant AI voice cloning that sounds like you—and stays saved when you need it

Audo is an online AI voice cloner and voice cloning software tool that clones a voice from a single 5-second to 3-minute audio or video recording and reads written scripts up to 8,000 characters (about eight minutes of speech) as a 192 kbps MP3, powered by Fish Drama 3. You can test instant voice cloning free on scripts up to 300 characters 3 times a day with no sign-up, or sign in to save reusable voices across web, REST API, and AI assistants.

When you clone your voice for podcast pickups, course updates, audiobook chapters, or YouTube voiceovers, a clean conversational recording matters far more than a long sample. Speaking naturally for 15 to 30 seconds in a quiet room captures your pitch, breath, and cadence better than several minutes in an echoey room. Because generative speech enhancers rebuild vocal formants and make clones sound generic, Audo's built-in sample cleanup uses non-generative noise removal to strip steady fan hiss while keeping your real vocal character intact (read how to prepare a clean voice sample for cloning).

Layer your cloned MP3 over background music in Mix audio, swap it into an existing video with Replace audio in video, turn it into a social waveform video in MP3 to MP4, or automate voice generation from scripts and AI agents via REST (clone_voice and /api/v1/voices) and MCP (list_voices and clone_voice).

Ways to clone, generate, or prepare a voice on Audo AI

| Workflow on Audo AI | Input required | Model & languages | Free tier & pay-as-you-go rate | Source |
|---|---|---|---|---|
| Instant voice cloning (/voice-cloning) | 5 s to 3 min audio/video clip (10–30 s recommended) + script up to 8,000 chars | Fish Drama 3 · 30+ languages + [whispers], [small laugh] tags | Free up to 300 chars 3x/day (no sign-up); then $0.12 per 1,000 chars, no subscription | https://audo.ai/voice-cloning?utm_source=markdown&utm_medium=agent |
| Saved cloned voices (Web, REST API & MCP) | One saved voice sample reused by voice_id across scripts | Fish Drama 3 · up to 3 saved voices free, 10 on paid accounts | Saving a voice is free; speech generation is $0.12 per 1,000 chars | https://audo.ai/docs/api?utm_source=markdown&utm_medium=agent |
| Preset AI text to speech (/text-to-speech) | Written script up to 8,000 chars (no voice sample needed) | Eleven v4, Eleven v4 Turbo (70+ languages) & Gemini 3.8 (130+ languages) | Free up to 300 chars 3x/day; then $0.05 to $0.20 per 1,000 chars | https://audo.ai/text-to-speech?utm_source=markdown&utm_medium=agent |
| Voice sample clean-up (/noise-removal) | Noisy microphone, phone, or video recording before cloning | Non-generative speech noise filter (any language, preserves vocal timbre) | First minute free 3x/day; then $0.05/min ($3.00/hr) | https://audo.ai/guides/clean-voice-sample-for-cloning?utm_source=markdown&utm_medium=agent |

Tool: Voice cloning (`clone_voice`). Price: 12 credits per 1,000 characters — $0.12 per 1,000 characters, pay-as-you-go with no monthly subscription — per job’s measured quantity, charged only when a job succeeds.

## Free API

```bash
curl -F file=@voice-sample.mp3 -F 'text=Welcome back to this week'\''s episode.' -F consent=true https://audo.ai/api/try/voice-cloning -D response.headers -o response.body
```

Read response.headers and response.body before naming media. Handle 202, 200 JSON and errors with the complete response client at https://audo.ai/docs/api?utm_source=markdown&utm_medium=agent#responses.

Free for scripts up to 300 characters, 3 times a day per person, with no key or account. Plain UTF-8 text or TXT uploads up to 8,000 characters (2 MB). Free results last one hour. The response is the result itself; a job that runs longer than about 30 seconds returns 202 with a link to check on it. With an API key (`-H "Authorization: Bearer $AUDO_API_KEY"`), omitting free runs the whole file with credits if the balance covers it, otherwise the free portion if quota remains. Set free=true to prohibit spending; for paid whole files use free=false with an approved positive max_credits. A free result cut short comes with finishUrl (and the Audo-Finish-Url header when the response is the file itself): give it to the person, who pays there (from $5, no sign-up), and the whole file runs; keep checking the status link, whose finishedBy names the whole-file job and its own status link. Details: https://audo.ai/docs/api.md

## Options

Form fields with the same names as the MCP tool's inputs:

- `text`: The script to speak in the cloned voice, up to 8,000 characters. Supports inline stage directions such as [whispers], [small laugh], [sighs], or [excited]. Pass text or text_file_id, not both.
- `text_file_id`: 
- `voice_id`: A saved cloned voice ID from list_voices (or POST /api/v1/voices). Pass voice_id or file_id, not both.
- `consent`: Required true when passing file_id: confirms you own this voice or have permission from the speaker to generate speech with it.
- `clean_noise`: Removes steady background hiss and room hum from file_id before cloning. Default true.
- `language`: Optional BCP-47 language tag for the script, such as en, es, or ja. Default: detected automatically from the script.

## Languages

30+ languages across cross-lingual speech synthesis (detected automatically from the script text, or set with `language`): `en` English, `es` Spanish, `fr` French, `de` German, `it` Italian, `nl` Dutch, `pt` Portuguese, `ja` Japanese, `ko` Korean, `zh` Chinese (Mandarin), `yue` Cantonese, `hi` Hindi, `ar` Arabic, `ru` Russian, `pl` Polish, `tr` Turkish, `sv` Swedish, `da` Danish, `nb` Norwegian, `fi` Finnish, `cs` Czech, `el` Greek, `he` Hebrew, `hu` Hungarian, `ro` Romanian, `uk` Ukrainian, `vi` Vietnamese, `id` Indonesian, `th` Thai.

## From an assistant (MCP)

Add https://audo.ai/mcp and sign in (setup for each assistant: https://audo.ai/setup.md). The tool is `clone_voice`:

> Clones a speaker's voice from a short audio or video sample (file_id with consent: true) or a saved account voice (voice_id from list_voices), and reads a script aloud as an MP3 up to 8,000 characters across 30+ languages. Direct delivery with bracketed tags in the script such as [whispers], [small laugh], [sighs], or [excited]. Costs 12 credits per 1,000 characters, at least 1, charged only if the job succeeds; free up to 300 characters, 3 times a day. Returns the spoken MP3 with download links, or job_id for get_job if it takes longer.

Example: "Speak these release notes in my saved cloned voice."

## Questions

**Is Audo's AI voice cloner free?** Audo generates cloned speech up to 300 characters free (about 20 seconds of speech), 3 times a day, with no sign-up required. Longer scripts up to 8,000 characters cost $0.12 per 1,000 characters (about a minute of speech), paid once from $5 with no subscription—see pricing. Free accounts can save up to 3 cloned voices for reuse, and paid accounts can save up to 10.

**How do I clone my voice from an audio or video file?** Upload 10 to 30 seconds of natural, single-speaker speech at a normal pace without background music or overlapping voices, or record straight in your browser on Audo. Audo's voice cloner accepts WAV, MP3, M4A, FLAC, OGG, MP4, and MOV files from 5 seconds to 3 minutes long. Leaving Remove background noise switched on runs Audo's non-generative noise filter so steady room hum is removed without flattening your vocal timbre, or you can clean a noisy clip first in Remove noise.

**Can I direct emotion, whispers, and laughs in my cloned voice?** Yes. On Audo, you can place bracketed acting directions right inside your script—such as [whispers], [small laugh], [sighs], or [excited]—to guide how Fish Drama 3 delivers each line while keeping your cloned voice's natural tone and cadence.

**What languages can a cloned voice speak on Audo?** Audo's cross-lingual voice cloning speaks 30+ languages from a single voice sample—including English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Russian, Turkish, Arabic, Hindi, Japanese, Korean, Mandarin Chinese, Indonesian, Vietnamese, and Thai—detected automatically from the script text.

**Do I have to upload my voice sample every time I generate speech?** No. Sign in on Audo and save your voice on the page to keep it in your account (up to 3 saved voices free, or 10 after any purchase). You can pick a saved voice in one click on the web, pass its voice_id to the REST API, or ask Claude, ChatGPT, or Cursor to read a script in your saved voice through Audo's MCP server, and delete saved voices anytime.

**What is the difference between Audo's Voice cloning and Text to speech?** Audo's Voice cloning clones a real speaker from a 5-second to 3-minute reference recording (using Fish Drama 3 at $0.12 per 1,000 characters) so the generated MP3 sounds like you. Text to speech uses built-in studio voices from Eleven v4, Eleven v4 Turbo, and Gemini 3.8 ($0.05 to $0.20 per 1,000 characters) when you want a ready-made narrator without uploading a voice sample.

## More

- Every tool: https://audo.ai/llms.txt
- OpenAPI: https://audo.ai/api/openapi.json
- This page for people: https://audo.ai/voice-cloning?utm_source=markdown&utm_medium=agent