Frost

Open Source Alternative to Vercel, Netlify, Railway, Render and Neon.

Build, deploy, and run from one open-source platform. Your apps run on your server.

Demo: demo.frost.build

Why Frost

• Open source runtime. No platform markup.
• Git push deploy workflow.
• Docker-native. No Kubernetes setup.
• AI-friendly deploy model with clear logs.

What you get

• Unlimited services, deploys, and seats
• Automatic SSL (Let's Encrypt)
• Custom domains and redirects
• PR preview environments
• Zero-downtime deploys and instant rollbacks
• Health checks, resource limits, persistent volumes
• Managed object storage with S3-compatible buckets and access keys
• REST API and MCP support for AI agents

Core model

• Project: container for related services with shared env vars
• Service: deployable unit from repo or prebuilt image
• Object Storage: S3-compatible API endpoint with buckets and scoped keys
• Domain: custom domain attached to a service (proxy or redirect)
• Deployment: immutable deploy record with logs and status

Services in one project share a Docker network. Use service name as hostname.

Object storage exposes one user-facing S3 API endpoint. Bucket and access-key administration lives inside Frost, so users do not need a second public console domain for the storage engine.

Install

With AI agent:

• Give your agent https://frost.build/install.md

Manual:

curl -fsSL https://frost.build/install.sh | sudo bash

Need VPS provisioning help: INSTALL.md

One simple deploy example

Create a project, deploy, then check status:

PROJECT_ID=$(curl -s -X POST http://localhost:3000/api/projects \
  -H "Content-Type: application/json" \
  -d '{"name":"demo","repo_url":"./test/fixtures/simple-node","port":3000}' | jq -r .id)

DEPLOYMENT_ID=$(curl -s -X POST "http://localhost:3000/api/projects/$PROJECT_ID/deploy" | jq -r '.deploymentIds[0]')

curl "http://localhost:3000/api/deployments/$DEPLOYMENT_ID"

Docs

• Product docs: https://frost.build/docs
• API docs: https://<your-frost-host>/api/docs
• GitHub repo: https://github.com/elitan/frost

Tech stack

• Bun + Next.js
• SQLite + Kysely
• Tailwind + shadcn/ui
• Docker

Local development

bun install
bun run dev

Open http://localhost:3000

Common test commands:

bun run e2e:local
bun run e2e:smoke
bun run e2e:changed:fast

Useful E2E knobs:

E2E_GROUPS='01-basic,29-mcp' bun run e2e:local
E2E_GROUP_GLOB='group-2*.sh' bun run e2e:local
E2E_BATCH_SIZE=4 E2E_START_STAGGER_SEC=1 bun run e2e:local

Image pull retries are built in: FROST_IMAGE_PULL_RETRIES, FROST_IMAGE_PULL_BACKOFF_MS, FROST_IMAGE_PULL_MAX_BACKOFF_MS.

Requirements

• VPS/server with Docker
• Ubuntu 20.04+ recommended

License

MIT