website/src/content/posts/2026-06-17-introducing-rivet-compute/page.mdx
Today we're releasing Rivet Compute: a serverless platform dedicated to hosting your actors.
Until now, running an actors backend meant bringing your own infrastructure. You either self-hosted the engine or deployed to a third-party platform like Vercel, and you owned the provisioning, scaling, and operations that came with it. Rivet Compute changes that. It's the first time you can deploy your actors directly onto infrastructure we run, with a single command and nothing to manage.
Point the Rivet CLI at your project and deploy:
npx @rivetkit/cli deploy
The CLI resolves your project from the token, builds and pushes your Docker image to Rivet's built-in registry, provisions the managed pool, and prints your deployment URL once the pool is ready. There's no separate registry to configure, no cluster to stand up, and no infrastructure to wire together. Your backend is live as soon as the status reaches Ready.
Rivet Compute isn't a generic container host that you have to adapt your app to. It's purpose-built for actors, so the runtime, networking, and actor lifecycle all work the way your app expects out of the box.
After your first local deploy, scaffold a GitHub Actions workflow that deploys on every push:
npx @rivetkit/cli setup-ci
This writes .github/workflows/rivet-deploy.yml, which creates production and pull-request namespaces, posts preview links on every PR, and cleans up preview namespaces when pull requests close. Set your token as a repository secret and every branch ships automatically:
gh secret set RIVET_CLOUD_TOKEN
Rivet Compute meters usage per minute and bills monthly. You're only charged while your instances are actually running, based on the memory and vCPU they use. There's no preview list and no waitlist. It's available today. See the pricing page for the current rates.