I built this free AI voice generator to turn text into speech without an account. Paste your text, choose a model and an English voice, and the complete audio is generated on your device.
You can choose between Phonon, with 25 English and 12 character voices, or Kokoro, with 28 American and British English voices. Both run in your browser, trying available WebGPU acceleration with a CPU fallback. Generation time depends on your device.
[0.5s] in the text for half a second of silence, or [1s] for a full second.Every clip reads the same line: “Some mornings, the coffee tastes better than usual — and I never quite know why.”
Voice quality varies by model and voice. I update the generator as their developers release new versions.
Both models produce 24 kHz speech. WAV is uncompressed 16-bit mono (about 2.9 MB per minute). MP3 uses 128 kbps CBR, about 960 KB per minute.
Yes, but smaller sections are more practical. The 50,000-character limit prevents unusually large jobs; it does not guarantee that a single job of that size will finish on every device.
Phonon offers 25 regular English voices and 12 English character voices. Kokoro offers 20 American English and 8 British English voices. Other languages and voice cloning are not included.
Phonon requires WebAssembly SIMD and module workers. WebGPU is optional: if it cannot start, the same model runs on the CPU. This page uses one CPU thread and does not require cross-origin isolation. Kokoro also supports a WebAssembly CPU fallback. Performance and completion depend on your browser and device.
Phonon's pinned q8 weights are 64,512,352 bytes (61.5 MiB), plus roughly 4.2 MiB of WebAssembly, its tokenizer, config and selected voice. These are file sizes, not a measured total browser transfer. Each model loads only when you generate with it. Kokoro's WebGPU model is roughly 326 MB. The WASM model is roughly 92 MB, plus its runtime and other assets. A failed WebGPU attempt may download both models. Your browser can cache these files, but it may evict or revalidate them. Clear this site's stored data in your browser settings to remove them.
Speech synthesis runs on your device. The generator does not upload the text you enter or the audio it creates. It still loads Phonon's runtime from this site, and downloads model files and voices from Hugging Face. Kokoro and the MP3 encoder also load scripts from jsDelivr. Those providers receive ordinary request metadata, including your IP address, browser, requested URL, and the time of the request. In production, Fathom records page views and the fixedVoice generated event. It does not receive your text or audio from this generator.
I place no additional restrictions on the audio you generate with this tool, whether for personal or commercial projects. You are responsible for checking that you have the rights to the source text, that its license permits your intended use, and that you comply with any other rights that apply. I do not check those rights or take responsibility for how you use the audio.