Runs in the browser · Speech to text
Run Whisper Base in your browser
Whisper Base downloads 139.3 MB over WebGPU (fp16), or 73.3 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-base
- Parameters
- 72.6M
- Pipeline task
- automatic-speech-recognition
Reading
The WASM build downloads smaller here: 73.3 MB (uint8) against 139.3 MB (fp16) for WebGPU. The CPU-fallback quant compresses tighter than the GPU pick for this model.
The quant ladder spans 3.8x: 73.3 MB (uint8) to 277.5 MB (fp32).
Will it run in your browser?
Checking for WebGPU support…
All measured variants
| Variant | Size |
|---|---|
| uint8 WASM pick | 73.3 MB |
| bnb4 | 133.2 MB |
| q4 | 135.8 MB |
| fp16 WebGPU pick | 139.3 MB |
| q8 | 146.7 MB |
| fp32 | 277.5 MB |
Sizes measured from the HuggingFace API file tree for onnx-community/whisper-base, not estimated.
Use it with Transformers.js
import { pipeline } from "@huggingface/transformers";
const pipe = await pipeline("automatic-speech-recognition", "onnx-community/whisper-base", {
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 Base, 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 Base?
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.