The backend for AI apps.

The one devs, CFOs and security teams love. Open a project — it's all there, today. Why devs love it Secrets, crons, keys — wired the moment a project opens. Swap the model or the machine under a live session: next month's model is a config change, not a rewrite. Why CFOs love it Every run attributed — per product, per user. Caps on by default: one runaway agent can't write the invoice. Most runs ride subscriptions you already pay. The bill gets boring. Why security teams love it Keys stay in your vault — values never transit cerver. Client-safe publishable keys for the browser. Every run leaves a transcript: an audit trail by default.

no card · $5 free tier

01 — At scale

Run thousands of agents in parallel. See every one.

Fan out to thousands of sessions at once. Click any agent and watch its session — the chat, the model, the compute, the cost.

02 — The mix

Stop running everything on the most expensive model.

Most fleets run 100% frontier by reflex. Your work isn't 100% frontier-shaped: about a quarter earns the best model, the rest doesn't notice the difference. A routing policy — rules you write, or one line of auto — makes the mix deliberate.

Today · all frontier

10,000 sessions$4,000 / mo

Right-sized · 25 / 35 / 40

same 10,000 sessions$1,540 / mo · −62%
Browse routing policies → What it saves you

03 — Swap underneath

Keep the transcript. Swap the agent and the machine under it.

Same stage, same actor. The harness and the compute are props you can change mid-run — the transcript, the tools, and the identity stay bound.

transcript · sess_42live · 14 turns
harnessclaude
computevercel

04 — Compare

Run the same task on two agents. Stay on whichever wins.

The comparison runs itself, in one session. A real diff — not vibes.

$ cerver run --compare claude codex "add idempotency to the webhook"

  claude / sonnet   passes · retries on 5xx
  codex  / gpt-5   passes · cleaner backoff   ← better

→ you're now on codex for this task.

05 — Anywhere

Local and/or cloud.

Hosted model sessions work instantly. When the session needs your repo, terminal, tools, or CLI agents, attach the relay and keep the same session boundary.

Models
Claude Opus 4.8
GPT-5
Grok 4
Gemma 4
Runtimes
Claude Code
Codex CLI
OpenAI SDK
xAI
Compute
Vercel
E2B
Cloudflare
Modal
your machine

New · Benchmark 2026

The State of AI-Coding Spend 2026

What teams really spend, how much is avoidable, and the model + tool prices behind it — with sources.

Read the report →
$500–2kper dev / month, heavy teams
30–40%of that is avoidable
5–8×cheaper to route routine work

06 — Start

Start in under a minute.

$ curl -fsSL https://cerver.ai/install.sh | bash
$5 free tier · no card