WebAssembly
Web Development
Performance
AI
Python
Blog AI Assistant
Ask anything about this article
Hi! I've read this article.
What would you like to know?
Ask anything about this article
Hi! I've read this article.
What would you like to know?
@farhan
WebAssembly (Wasm) has evolved from a niche technology to a fundamental building block of the modern web. In 2025, you can run Python, execute ML models, and build full desktop-class applications entirely in the browser.
html<!DOCTYPE html>
<html>
<head>
<script src="https://cdn.jsdelivr.net/pyodide/v0.25.0/full/pyodide.js"></script>
</head>
<body>
<script>
async function main() {
const pyodide = await loadPyodide();
// Install packages
await pyodide.loadPackage(['numpy', 'pandas', 'scikit-learn']);
// Run Python code
const result = pyodide.runPython(`
import numpy as np
from sklearn.linear_model import LinearRegression
X = np.array([[1], [2], [3], [4], [5]])
y = np.array([2, 4, 5, 4, 5])
model = LinearRegression()
model.fit(X, y)
prediction = model.predict([[6]])
f"Prediction for x=6: {prediction[0]:.2f}"
`);
console.log(result); // "Prediction for x=6: 5.40"
}
main();
</script>
</body>
</html>
javascriptimport * as ort from 'onnxruntime-web';
// Load a pre-trained model
const session = await ort.InferenceSession.create(
'./model.onnx',
{ executionProviders: ['wasm'] }
);
// Run inference
const inputTensor = new ort.Tensor('float32', imageData, [1, 3, 224, 224]);
const results = await session.run({ input: inputTensor });
const predictions = results.output.data;
javascriptimport { pipeline } from '@xenova/transformers';
// Sentiment analysis in the browser
const classifier = await pipeline('sentiment-analysis');
const result = await classifier('I love WebAssembly!');
// [{ label: 'POSITIVE', score: 0.9998 }]
// Text generation
const generator = await pipeline('text-generation', 'Xenova/gpt2');
const output = await generator('WebAssembly enables', { max_length: 50 });
| Task | JavaScript | Wasm | Native | Wasm vs JS |
|---|---|---|---|---|
| Matrix multiply (1000x1000) | 1200ms | 180ms | 150ms | 6.7x faster |
| Image resize (4K) | 450ms | 85ms | 60ms | 5.3x faster |
| JSON parse (10MB) | 380ms | 120ms | 90ms | 3.2x faster |
| SHA-256 hash (1GB) | 8500ms | 1200ms | 900ms | 7.1x faster |
| ML inference (ResNet) | N/A | 200ms | 45ms | Works! |
WASI (WebAssembly System Interface) makes Wasm a universal runtime:
bash# Compile Rust to WASI
cargo build --target wasm32-wasi
# Run anywhere: browser, server, edge, IoT
wasmtime ./target/wasm32-wasi/release/myapp.wasm
# Or with Node.js
node --experimental-wasi-unstable-preview1 run.js
WebAssembly is becoming the universal binary format:
The browser is becoming the new operating system, and WebAssembly is its machine code.
