bring your own plan

You already pay for the model. Stop paying to use it.

Cortev runs your coding agents in the cloud on the AI subscription you already have — Claude, ChatGPT, Gemini. Your account does the work, so the token meter never starts. We charge for the cloud, not for what it thinks.

your account, your credentials no API keys to buy revoke any time

Invoice · metered API this month
checkout refactor · storefront$412.80
dependency sweep · 38 repos$918.44
overnight test backfill$604.11
security pass · every line$311.09
…and the runs still going+
amount due $2,246.44$2,451.09$2,663.77$2,847.19 ▲ climbing
on Cortev
Your plan · already paid unchanged
Claude Max · your accountalready yours
the same four jobs aboveincluded
extra to run them $0.00

Illustrative figures — what you'd actually save depends on your usage. The point isn't the number, it's which column it lands in.

The invoice

Agent work is metered on an API and flat on a plan.

That difference is the entire product. Cortev sells you the cloud and the harness — never the tokens.

The usual way Per tokenmetered API credits

Every run bills again, and the meter keeps running while an agent sits there thinking. The longer the job, the more it costs to do the work that matters most.

The Cortev way Your planflat · already paid for

The same work runs under the subscription sitting in your account right now. Nothing new to buy, nothing metered, nothing to reconcile at the end of the month.

What you pay us The cloudhosting + orchestration

Runners that stay up, the harness that routes and supervises the work, the vault, and one screen to watch it all. Priced like infrastructure, because that's what it is.

How much you save depends on how hard you push it — the gap is widest for always-on, multi-agent work. Your plan's own fair-use limits still apply.

How it works

Three steps, then close the laptop.

01

Connect your plan.

Sign in with your own Claude, ChatGPT, Gemini or Copilot account. The credentials land in your vault, scoped to your workspace, and you can pull them back out whenever you want.

02

Point it at your work.

Repos, accounts, servers, DNS — mapped in about sixty seconds and read-only until you say otherwise, including the projects you'd forgotten were still running.

03

Close the laptop.

Agents run on our runners, not your machine. They keep working while you sleep and report back where you can see them — including from your phone.

What you get

A real cloud workspace, running on your own account.

cortev cloud · your workspace 3 agents running
signed in with
Claude Max connected
account · you@yourcompany.com
billing · your existing plan
or connect ChatGPT, Gemini,
Copilot — or your own API key
running on cortev cloud runners
storefront · fix the checkout race condition 1m 12s
getaway · add email + password auth 44s
market · upgrade Stripe checkout shipped
your laptop · closed metered API spend · $0.00
Hosted

The cloud does the sitting.

Long jobs don't belong on a laptop that sleeps, roams between networks and runs out of battery. Cortev's runners stay up, so a two-hour refactor finishes whether you're at the desk or on a plane — and the work is still there when you get back.

uptime · this workspace
41d 06:12:58
runner-2 · storefront working
watcher · 38 repos scanning
Your account

Your plan. Your credentials. Your call.

There's no shared pool and nothing resold. You authenticate as yourself, the secrets sit in your own vault, and one click cuts the connection. It's your subscription doing your work — we just gave it somewhere to run.

your vault · scoped to you
🔒 claude · oauth session yours
🔒 github · repo scope yours
🔒 aws · read-only yours
revoke any row → agents stop using it on the next run
One harness

Any model, without re-plumbing anything.

Connect more than one plan and Cortev routes each task to whichever model you've made available — the strong one for the gnarly refactor, the cheap fast one for the sweep. Hit a limit on one and the work moves rather than stopping.

orchestrator · routing
Opus Sonnet GPT Gemini
refactor checkout → Opus · your plan
metered API credits used → none
Straight answers

The questions everyone asks first.

Whose account does the work actually run under?

Yours. You sign in with your own plan inside your own workspace, and every request goes out as you. Cortev doesn't pool accounts, doesn't share seats and doesn't resell anyone's access — there's nothing to resell, because the account was always yours.

Then what am I paying Cortev for?

The cloud runners that stay up, the harness that routes and supervises the work, the vault, the watchers and the one screen that shows you all of it. Infrastructure and orchestration — not tokens, and not a markup on someone else's model.

What happens when I hit my plan's limits?

Your provider's fair-use limits apply exactly as they do on your own machine. Cortev shows you where you stand and can hand the task to another model you've connected, so a limit slows you down instead of silently turning into a bill.

Can I still pay per token if I want to?

Yes. Bring an API key and Cortev will use it — some teams prefer metered billing for workloads they want charged to a project. You can run a key and a subscription side by side and let the orchestrator choose.

Case studies

Who runs this way.

get started

Put the plan you already pay for to work.

Connect your account, point Cortev at your repos, and let it run while you're somewhere else. No credits to buy to find out whether it works.