Skip to content

Runs in the browser · Speech to text

Run Whisper Tiny in your browser

Whisper Tiny downloads 72.6 MB over WebGPU (fp16), or 39.0 MB over WebAssembly (uint8), to transcribe speech to text entirely in the tab with Transformers.js. No install, no server.

Speech to text · onnx-community/whisper-tiny

WebGPU download
72.6 MB
fp16
WASM download
39.0 MB
uint8
Parameters
37.8M
Pipeline task
automatic-speech-recognition

Reading

The WASM build downloads smaller here: 39.0 MB (uint8) against 72.6 MB (fp16) for WebGPU. The CPU-fallback quant compresses tighter than the GPU pick for this model.

The second-smallest speech to text download in the catalog (72.6 MB).

The quant ladder spans 3.7x: 39.0 MB (uint8) to 144.4 MB (fp32).

Run it in your browser

Live run this browser

Checking what this browser can run…

All measured variants

Quant Download size
Variant Size
uint8 WASM pick 39.0 MB
fp16 WebGPU pick 72.6 MB
q8 77.9 MB
bnb4 90.3 MB
q4 91.3 MB
fp32 144.4 MB

Sizes measured from the HuggingFace API file tree for onnx-community/whisper-tiny, not estimated.

Use it with Transformers.js

import { pipeline } from "@huggingface/transformers";

const pipe = await pipeline("automatic-speech-recognition", "onnx-community/whisper-tiny", {
  device: "webgpu", // usually falls back to "wasm"; wrap in try/catch for production
  dtype: "fp16", // use "uint8" for the WASM build
});

const result = await pipe(/* your input */);

Requires npm install @huggingface/transformers (or the CDN build).

Sources

Same job, different size

FAQ

WebGPU or WASM for Whisper Tiny, in practice?

Chrome, Edge, and Safari 26+ run the WebGPU build (fp16) on the GPU. Firefox has WebGPU on Windows and Apple Silicon Macs but not everywhere yet. Any browser without a WebGPU adapter falls back automatically to the WebAssembly build (uint8) on CPU: same model, slower to load, slower to run.

What does fp16 mean for Whisper Tiny?

The WebGPU build here uses fp16: 16-bit floating point (half precision): smaller than fp32 with effectively no quality loss. The WASM fallback uses uint8: 8-bit integers: about a quarter the size of fp32, with a smaller quality trade than 4-bit.

Sizes measured 2026-08-01 from the HuggingFace API. Last validated 2026-08-03. See all browser models or the methodology.