The control plane for agent compute

Cerver routes agent work to the right computer. Your app opens a session, Cerver picks a provider — your own machine via Cerver Connect or a cloud sandbox — and Cerver Runtime runs the work wherever it lands.

Read docs See the model

Usage

your_app.ts
const session = await fetch("/gateway/sessions", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({
    task: "Boot a preview and run tests",
    requirements: { runtime: "node", timeout_minutes: 20 },
    policy: { mode: "balanced" }
  })
})

await fetch(`/gateway/sessions/${sessionId}/run`, {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({
    code: "npm test",
    timeout: 600
  })
})

session result
{
  "session_id": "sess_123",
  "provider": "vercel",
  "compute_id": "cmp_123",
  "stdout": "test suite passed",
  "metrics": {
    "provision_time_ms": 812,
    "last_exec_latency_ms": 1432,
    "cost_estimate_usd": 0.01
  }
}

Four products, one platform

Cerver

The control plane. Sessions, routing, metrics, and policy. Your app talks to Cerver — never directly to providers.

Cerver Runtime

A portable compute engine — git, worktrees, dev servers, CLI execution. Runs the same on your laptop, Vercel, E2B, or any future provider.

Cerver Connect

The transport layer for private compute. Your machine opens a persistent WebSocket to Cerver and becomes a routable provider — no inbound ports needed.

Cerver SDKs

Client libraries so your app can open sessions, run code, and stream output in a few lines of Python or TypeScript.

Visible by default

Track startup, latency, stream timing, uptime, engagement, and estimated cost per session.

Bring your own providers

Add a new backend by implementing the compute adapter contract. Cerver Runtime runs on it automatically.

Startup profiles

Warm local compute fast

Warm remote compute balanced

Fresh remote compute cold start

Different computers wake up differently. Cerver keeps the session API stable and reports the startup profile back as metrics.

How it works

Register compute

Add cloud providers or connect your own machines via Cerver Connect.

Open a session

Your app describes what it needs — runtime, timeout, policy. Cerver picks the right provider.

Run and stream

Cerver Runtime executes the work — same engine whether it lands on your laptop or a cloud sandbox.

See what happened

Startup time, latency, uptime, and cost — all reported back through the session.