Frequently asked questions
Quick answers about how Whisperly's free AI voice generator and on-device text to speech works — no sign-up, AI voices, offline use, and more.
Is Whisperly free?
Yes — Whisperly is a completely free AI voice generator, with no account, no sign-up and no API keys. The speech is generated on your own device, so there are no per-character costs, quotas or rate limits.
How does Whisperly make money?
Right now it doesn't — Whisperly is free, with nothing to buy and no ads. The on-device text to speech that the app is built around will stay free. We're planning an optional Cloud Sync feature that keeps your sessions in sync across your devices; that is the kind of extra that could support the project later, while the core app remains free to use.
Is Whisperly an AI voice generator?
Yes. Whisperly is a free AI voice generator and text-to-speech app: it turns your text into natural AI voices using neural models (Kokoro and Piper) that run right in your browser. Unlike most AI voice tools it runs on-device, which is what makes it free, private and able to work offline.
Is there really no sign-up or character limit?
Correct. There is no sign-up, no login and no API key, and because the AI voices run on your own device there is no character limit — you can convert anything from a single sentence to a whole article. It is genuinely free, unlimited AI text to speech.
Does my text leave my device?
No. Whisperly runs entirely in your browser, so your text and the audio it generates never leave your device and are never sent to a server. There is no backend that could see them.
Do I need to install anything?
No. Whisperly runs in any modern browser, with nothing to download or install. The first time you use a voice, its model is fetched once and cached in your browser; after that it loads instantly. If you like, you can also install Whisperly to your device from your browser, so it opens in its own window like a regular app — but it works exactly the same either way.
Why does the first conversion take a moment?
The first time you use a voice, Whisperly downloads its model — that is what makes on-device, offline synthesis possible. The download happens once and is then cached in your browser, so every conversion after it is fast.
Does it work offline?
Yes, once a voice model has downloaded. The model is cached in your browser, so you can convert text to speech with no network connection at all.
How is Whisperly different from cloud text-to-speech tools?
Most text-to-speech tools run in the cloud: you send them your text, they charge per character, and they need an internet connection and usually an account. Whisperly runs the voices on your own device instead, which is what makes it free, private and able to work offline, with no sign-up.
Which voices and languages are available?
Whisperly ships the multilingual Kokoro model — dozens of voices across English, Chinese, German, Spanish, French, Italian, Portuguese, Japanese and Hindi — plus dedicated Piper voices such as Amy (US English) and Thorsten (German). You can adjust the speaking speed for any voice.
How does in-browser text to speech work?
Whisperly runs the sherpa-onnx neural TTS engine, compiled to WebAssembly. The voice model runs directly on your CPU inside the browser tab, so synthesis happens locally instead of on a remote server.
Can I download the generated audio?
Yes. Whisperly turns your text into audio you can play back and download as an MP3 to use wherever you like.
Can I use Whisperly as free TTS for streaming?
Yes. Whisperly works well as free text to speech for streamers — generate the AI voice clip, download the MP3, and play it through your streaming setup. There is no sign-up and no character limit, and it keeps working offline once a voice has downloaded.
Can I use the generated audio commercially?
Yes. The voice models Whisperly ships are released under permissive open-source licenses (MIT, Apache-2.0, CC0 and CC BY 4.0), so the audio you generate is yours to use — including in videos and other commercial projects. Each model's license and any required credit are shown in the app's voice settings.
Where are downloaded voices stored, and can I remove them?
Voice models are stored in your browser (IndexedDB), not on a server. You can see how much space they use and remove individual models — or clear them all — from the Model cache dialog inside the app.
Does Whisperly work on my phone?
Yes. Whisperly runs in modern mobile browsers and has a touch-friendly layout. Synthesis still happens on your device, so very large voice models are more demanding on phones than on a laptop, but smaller voices work well.
Which browsers are supported?
Whisperly works in current versions of Chrome, Edge, Firefox and Safari. It needs WebAssembly and a cross-origin isolated context (for SharedArrayBuffer), which modern browsers provide.
Is there a limit on how much text I can convert?
No. Because everything runs on your device, there are no usage limits. Longer texts just take a little more time to synthesize, and you can pause and resume long runs.