Universal floor
Try KittenTTS Nano free — unlimited, in your browser
KittenTTS Nano is a 15-million-parameter model whose full-precision browser bundle weighs about 57 MB — a fraction of any other voice model here. It is the model that works when nothing else does: no GPU, minimal RAM, quick download. Quality is basic but clear.
Echo Live High-quality · Any characterDownload the app
Runs entirely in your browser — nothing is uploaded.0 / 5000
Specifications
| Developer | KittenML |
|---|---|
| Parameters | 15 million |
| Download size | 57 MB (fp32) |
| Output | 24.0 kHz mono |
| Voices | 8 |
| Languages | English (US) |
| Voice cloning | No |
| Streaming | No |
| License | Apache-2.0 |
| Speed | Faster than real-time on virtually any device, including old phones. |
| Runtime | WebAssembly (any CPU) |
Source: KittenML/kitten-tts-nano-0.8
Strengths
- +57 MB download — loads fast on any connection
- +Runs on effectively 100% of devices, single-threaded CPU included
- +Apache-2.0 end to end
Limitations
- −English only, 8 voices
- −Flattest prosody of the lineup
Voices (8)
Bella · en-USJasper · en-USLuna · en-USBruno · en-USRosie · en-USHugo · en-USKiki · en-USLeo · en-US
Compare
KittenTTS Nano — questions
Is KittenTTS Nano free to use?
Yes. The weights are Apache-2.0 and they run on your device.
What hardware does KittenTTS Nano need?
Any modern browser — it runs on CPU via WebAssembly, no GPU needed. The download is 57 MB, cached after first use.
Is my text private when using KittenTTS Nano here?
Yes. The model executes inside your browser tab — the text you synthesize is never transmitted to a server.
Can I use KittenTTS Nano output commercially?
Review the current model, voice, and source-dataset terms before commercial use. The catalog lists Apache-2.0 for this model.