Developer platform

Build and run AI apps on a global network.

Run open models on edge GPUs, route requests across providers, and add caching, rate limits, and observability to every AI call.

Get started free Talk to an expert

const answer = await env.AI.run('text-model', {
  messages: [{ role: 'user', content: 'Summarize this page' }],
});
return Response.json(answer);

🤖 AI Gateway & Inference

What you get

Serve popular open models for text, embeddings, and images from GPUs near users.

One endpoint for multiple model providers with fallbacks and retries.

Cache identical requests at the edge to cut latency and spend.

Store embeddings and run similarity search beside your functions.

Tokens, cost, and latency per model, per app, per user.

Filter prompts and responses with configurable safety policies.

⚡

Serve static and dynamic content from 180+ edge locations with tiered caching, smart routing, and instant purges.

Learn more

🛡

Unmetered DDoS mitigation, a managed WAF, bot management, and API protection in one policy engine.

Learn more

{}

Deploy serverless JavaScript, TypeScript, and WebAssembly to the edge with no cold starts and no regions to pick.

Learn more

Create a free account in under a minute. No credit card required.