v3.8 — streaming responses are GA

Functions that run where your users are

Write a handler, push it, and it's live in 42 regions in under ten seconds. No containers to size, no servers to patch — you pay only for the requests you serve.

Start free Read the quickstart

// api/hello.ts
export default async function handler(req) {
  const region = req.headers.get("x-lambda-region");
  return Response.json({
    hello: "world",
    region,
  });
}

$ lambda deploy
✓ Built in 412 ms
✓ Live on 42 regions
  https://hello-7f3a.lambda.run

<5 msmedian cold start

42edge regions

99.99%uptime, last 90 days

1Mfree requests / month

Everything a backend needs, nothing it doesn't

Zero-config deploys

Push from git or the CLI. Every branch gets its own preview URL.

Durable KV

Globally replicated key-value storage with read-your-writes consistency.

Cron & queues

Schedule jobs and process background work without a separate worker fleet.

Streaming

Send responses as they're generated — ideal for long-running and AI workloads.

Secrets

Encrypted environment variables, scoped per project and environment.

Observability

Logs, traces and per-route latency out of the box. Export to OpenTelemetry.