
Netlify
Where apps come to life.
5.0•62 reviews•3.2K followers
Where apps come to life.
5.0•62 reviews•3.2K followers
Build, Preview, and Deploy. All in One Place.
Go live in minutes with a unified development platform designed for speed and flexibility. Whether you prefer to build with AI or write your own code, you have everything you need to launch full-stack applications effortlessly.
This is the 18th launch from Netlify. View more

Netlify Server
Launched this week
Netlify Server gives your app a backend that stays alive while people are using it. Build realtime experiences, multiplayer games, and AI agents that need time to finish. Describe what you want and an agent builds it, or bring your Express, Fastify, Hono, or Koa code. It holds WebSockets open, keeps state in memory, and deploys with the rest of your project, with its own server for every preview. Scales with demand and down to zero when idle. Free during early access.






Free Options
Launch Team

AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted



Heck yeah, Netlify team! This is awesome. What a week of announcements for ya'll!
@thomascdaly Thank you so much. It's been an epic week 🎉
Let's go!
@nickytonline HELL YEAH!
Really excited about Netlify server! Awesome work and rebrand 👏
The long-running agent part is what caught me — I've had too many agent runs bump into function timeouts, so a server that just stays alive with state in memory is a big shift. How does the scale-to-zero side work with that — does active work keep a server alive, or can it spin down mid-task?
Excited to play around with Netlify Server. So much fun stuff to build with games and agents. I made a simple game a year ago, and just getting SSE or WebSockets set up was such a pain.
Giving each deploy preview its own backend is a useful detail. A reviewer can check the UI change alongside the server behavior it depends on, in the same preview.
@oleksii_sekundant agreed. Thanks for your support!
The long-running server model addresses a real gap for AI agents that need to maintain state across multiple tool calls. Being able to bring Express/Hono directly instead of refactoring for functions is huge - removes a whole category of onboarding friction for teams already invested in those frameworks.