voicedbyme Download

VoicedByMe / Use cases / Local ElevenLabs alternatives

Roundup · Voice cloning · Updated 24 October 2026

ElevenLabs, but on your own computer.

In 2026 you can clone a voice and generate speech entirely offline with open models such as Qwen3-TTS, Chatterbox and F5-TTS. The easiest ways in are desktop apps that bundle them: VoiceStudio and Voicebox are free and open source, and VoicedByMe is $30 once with nothing to set up. Check the model's licence before commercial use: F5-TTS's weights and XTTS-v2 are non-commercial.

Download for Windows Download for Mac Free for a day · then $30 once

Samples made offline with ready-made voices. All of them are AI-generated.

Why local

No credits, no uploads.

No meter running

A novel is about half a million characters. Locally, the hundredth hour costs the same as the first.

Your voice stays with you

Voice recordings are personal data. Locally, nobody else stores your voice or your scripts.

No account to lose

No credits that expire, no plan changes, no moderation queue.

Works offline

On a plane, behind a company firewall, or when the service is down.

The trade-offs

You need a reasonably recent graphics card or Apple silicon for good speed, open models speak fewer languages, and the most expressive cloud voices still lead on emotion.

The models

What the apps run underneath.

ModelFromCloningLanguagesLicence
Qwen3-TTS (0.6B, 1.7B)Alibaba's Qwen team, January 2026From about 3 seconds, with a transcript10, including English, Spanish, German, Chinese, JapaneseApache 2.0: commercial use allowed
Chatterbox (Multilingual, Turbo, Nano)Resemble AIFrom a reference clip23 (Multilingual); English (Turbo, Nano)MIT code; output carries Resemble's watermark
F5-TTSSWivid (research)From a short clipEnglish and Chinese, more from the communityCode MIT, weights CC-BY-NC: not for commercial use
XTTS-v2Coqui (closed in early 2024)From about 6 seconds17Coqui Public Model License: non-commercial
Kokoro-82MhexgradNo cloningAbout 8Apache 2.0

From each project's own page, October 2026. Licences can differ between code and weights: read both before commercial use.

The apps

Ways to run them without writing code.

VoiceStudio (free, open source)

An open-source studio for voice cloning, voice design, dubbing, dictation and audiobooks, with installers for Windows, macOS and Linux. Its default engine covers hundreds of languages, and you can switch engines. The app is AGPL-3.0; each model has its own licence, which the app shows but doesn't vouch for.

Voicebox (free, open source)

A desktop voice-cloning studio built originally on Qwen3-TTS, now with more engines to choose from. It clones from a few seconds of audio and runs on your machine, with builds for Mac and Windows; from source it needs a Python setup.

Pinokio with F5-TTS or AllTalk (free)

Pinokio is a free one-click launcher for open-source AI apps. It installs web interfaces such as F5-TTS's and AllTalk (which runs XTTS-v2, F5-TTS and others) without you typing commands. It still builds Python environments behind the scenes, takes several gigabytes, and is easiest with an NVIDIA graphics card.

Command line (free)

Each model above installs with pip and comes with examples. Most flexible, most work: Python, PyTorch for your graphics card, and FFmpeg.

VoicedByMe ($30 once)

A normal installer for Windows 10 and 11 and Macs with Apple silicon. Clone your voice from twenty seconds, or use 70 ready-made voices, then use it everywhere: read any app's text, make audiobooks with chapters, send voice notes, give AI agents a voice, or call its OpenAI-compatible API. Its output is watermarked and labelled as AI-made. It speaks English, Spanish, German, French, Portuguese, Italian, Japanese and Russian.

Compared

The apps side by side.

SetupPlatformsLanguagesBeyond cloningPrice
VoicedByMeAn installerWindows, Apple silicon Macs8, including English, Spanish, French, JapaneseReads any app, audiobooks, voice notes, agents, API$30 once
VoiceStudioAn installerWindows, Mac, LinuxHundreds (default engine)Dubbing, dictation, audiobooksFree (AGPL-3.0)
VoiceboxBuilds, or Python from sourceMac, WindowsDepends on the engineSeveral enginesFree, open source
Pinokio + F5-TTS or AllTalkOne click, then large downloadsWindows, Mac, LinuxDepends on the modelWeb interfacesFree
ElevenLabs (for reference)None, onlineAny browserDozensDubbing, voice library, agentsFrom $6 a month, credits

As of October 2026. Check each app's own site for its current features and prices.

Commercial use

The model's licence is the one that counts.

An app can be open source while the model inside it isn't free for commercial work. F5-TTS's code is MIT, but its trained weights are CC-BY-NC because of their training data. XTTS-v2's licence is non-commercial, and Coqui, which could have sold a commercial licence, closed in early 2024. Qwen3-TTS and Kokoro are Apache 2.0.

Whatever the licence, cloning someone else's voice needs their consent, and in the EU, AI audio has to be marked. More in is voice cloning legal?. VoicedByMe's licence covers commercial use of what you make with its voices or your own.

Questions

Can I run ElevenLabs locally?

No. ElevenLabs only runs online. To clone a voice on your own computer, use an open model like Qwen3-TTS or Chatterbox through an app such as VoiceStudio, Voicebox or VoicedByMe.

What's the best open-source voice cloning model in 2026?

For most people, Qwen3-TTS or Chatterbox: both clone from a short clip, speak several languages and have permissive licences. F5-TTS sounds excellent but its weights are non-commercial.

Can I use F5-TTS commercially?

The code is MIT, but the pretrained weights are CC-BY-NC, which doesn't allow commercial use. Train your own weights on licensed data, or pick a model with a commercial licence.

Is XTTS-v2 free for commercial use?

No. It's under the Coqui Public Model License, which is non-commercial, and Coqui closed in early 2024.

Do I need a graphics card?

For comfortable speed, a recent NVIDIA card or an Apple silicon Mac. Most models also run on the processor alone, several times slower: fine for saving files, slow for reading live.

Is there a local voice cloning app without Python?

Yes: VoicedByMe and VoiceStudio have normal installers, and Voicebox has builds for Mac and Windows.

Is it legal to clone a voice?

Your own, yes. Someone else's only with their permission. US and EU rules in 2026.

How much does it cost?

$30 once for the first 50 customers, then $50. No subscription. The full app is free for one day so you can try it first. A license is for one computer.

For Windows 10 and 11, and Macs with Apple silicon

Your voice, on your own computer.

$30once. Launch price for the first 50 customers, then $50.
One computer. No subscription.

Download for Windows Download for Mac Also for Windows 10 and 11 Also for Macs with Apple silicon

The full app, free for 24 hours. Purchases aren't refundable, so please try it first.