2 models · Text to speech · Transformers.js
Text to speech in the browser
2 models handle text to speech in the browser today, from 147.4 MB (Kokoro-82M) to 250.7 MB (Supertonic TTS 2), a 199.1 MB download at the midpoint. Every one runs with Transformers.js over WebGPU or WebAssembly: no server, no install, no account.
- Models
- 2
- Smallest
- 147.4 MB
- Median
- 199.1 MB
- Largest
- 250.7 MB
Every text to speech model, ranked by WebGPU size
| Model | WebGPU download | WASM download |
|---|---|---|
| Kokoro-82M 82M params | 147.4 MB q4f16 | 88.1 MB q8 |
| Supertonic TTS 2 | 250.7 MB fp32 | 250.7 MB fp32 |
Sizes measured from the HuggingFace API file tree for each model's repo, not estimated. Sorted smallest to largest by the WebGPU headline pick.
Size ladder
The WebGPU download spans 1.7x here: 147.4 MB (Kokoro-82M) to 250.7 MB (Supertonic TTS 2).
1 of 2 models (Kokoro-82M) download a smaller build over WebAssembly than WebGPU: the CPU-fallback quant compresses tighter than the GPU pick there.
None of the 2 models here fit under 100 MB over WebGPU.
A bytes-per-parameter comparison needs at least 3 models with a verified parameter count and a real spread between them; this group does not clear that bar yet.
Notable text to speech models
Other tasks
Or see every browser model grouped by task, or the main GGUF catalog for native runtimes outside the browser.
FAQ
Do all text to speech models here use the same WebGPU quant?
No. 1 use q4f16, 1 use fp32. q4f16 is the most common pick.
What would it cost to download every text to speech model here?
398.1 MB total over WebGPU across all 2 models, if you tried every one back to back. Most projects only need the single model that fits the job, not the whole set.
Sizes measured 2026-08-01 from the HuggingFace API. Last validated 2026-09-14. See all browser models or the methodology.