Codex hosting that remembers where it left off
Run OpenAI's Codex on a persistent, always-on Agent 37 Cloud instance from $3.44/mo. Provision it with one API call, hand it long jobs, and come back to a machine that still has the repo, the caches, and the context.
New workspaces get a $1 starter credit: first instance free, no card required.A coding agent on a disposable sandbox starts over every day
Ephemeral sandboxes are built for one-shot tasks: spin up, run, evaporate. Agentic coding work is not one-shot. It spans sessions, accumulates a working tree, installs toolchains, and picks up threads from yesterday. Hosting Codex on a persistent instance means the environment is part of the work, not a cost you pay again every morning.
The usual objection to persistence is price, because the box never turns off. Here that objection is $3.44 a month, metered per minute from a prepaid balance, on the same fleet that runs roughly 1,000 live instances today.
A persistent computer behind one API
Agent 37 Cloud provisions isolated, always-on instances from Docker images. Codex runs through the custom-image route, and the API treats it like any other workload.
Bring your image
There is no prebuilt Codex template, and we will not pretend otherwise. You package Codex in a public Docker image and register it as a custom template. That route is fully supported today.
One POST to provision
POST /v1/instances provisions a persistent computer running your image on always-on compute. Each instance is an isolated machine with its own disk.
A URL per instance
Every instance is reachable at https://{id}.agent37.app, with any port routable. No reverse proxy, no DNS, no certificates.
Streaming built in
POST /v1/responses on the instance URL runs a turn and streams text, reasoning, and tool activity over Server-Sent Events.
State that stays
Files, memory, and sessions persist until you delete the instance. Checkouts, dependencies, and half-finished branches are waiting next session.
Kernel-level isolation
Every instance runs under the gVisor runtime with a hard per-instance disk quota. Delete the instance anytime and billing stops.
What always-on actually costs
Estimated monthly cost of keeping a roughly 3 GB box on 24/7. Sandbox platforms are great at what they are for; a machine that never turns off is not what they are for.
The 2 vCPU / 4 GB shape is $4.94/mo, and disk expands at $0.09 per GB per month. Everything is metered per minute from a prepaid balance: no seats, no minimums, and billing stops the moment you delete an instance.
Three things, and one of them is free
Codex is a third-party tool: OpenAI builds it, you license it from them, and this page never pretends otherwise. Here is the full list.
Your OpenAI account
Codex is OpenAI's coding agent, not ours. You bring your own OpenAI account or API credentials, and OpenAI bills you for model usage directly. We host the computer it runs on.
A Docker image with Codex installed
Any public image works: install the Codex CLI, add the tooling your projects need, and register the image as a template. Instances provisioned from it are persistent from the first boot.
An Agent 37 Cloud workspace
Sign in, mint an API key, and provision. New workspaces get a $1 starter credit, roughly a week of the cheapest instance, so the first instance is free and needs no card.
Schedule it. Ship it under your own brand.
Scheduled and cron runs
Give an instance a cron schedule and it runs unattended: overnight batch fixes, dependency updates on a timer, a code-health report before standup. A persistent machine means every scheduled run starts warm.
White-label by default
Reselling a coding-agent service? You hold the API key, provision one instance per customer, and cap per-instance spend. Your customers see your brand, never Agent 37.
Codex hosting questions, answered
Is there a one-click Codex template?
No. Agent 37 Cloud has no prebuilt Codex template today. Codex runs as a custom template: you package OpenAI's Codex CLI in a public Docker image, register the image under a template name, and provision persistent instances from it with one POST. Everything else on this page, persistence, streaming, scheduled runs, and isolation, works with that route right now.
What does Codex hosting cost?
From $3.44/mo for an always-on 1 vCPU / 3 GB instance; the 2 vCPU / 4 GB shape is $4.94/mo. Extra disk is $0.09 per GB per month. Everything is metered per minute from a prepaid balance with no seats, and billing stops the moment you delete an instance. New workspaces get a $1 starter credit, so the first instance is free to try, no card required. Model usage is separate: you pay OpenAI directly through your own account.
Can Codex run 24/7 here?
Yes. Instances are persistent and always-on by design. A refactor that takes all night keeps running after you close your laptop, and cron-style scheduled runs execute unattended so the results are waiting in the morning.
Where does my code and data live?
On your instance's persistent disk, inside an isolated container running under the gVisor runtime with a hard per-instance disk quota. Files stay on the instance until you delete it, and you can delete anytime. Your OpenAI credentials live on your instance and go from there to OpenAI; we never need them.
Is Codex an Agent 37 product?
No. Codex is OpenAI's coding agent, and it needs your OpenAI account to run. Agent 37 Cloud is the infrastructure layer: a persistent, isolated, always-on computer with an API for provisioning, streaming, files, and schedules. The same platform also hosts the open-source Hermes agent and OpenClaw, and any public Docker image.
Can I host Anthropic's Claude Code the same way?
Yes, through the same custom-image route on the same API and pricing, and a one-click Claude Code template is coming soon. See the Claude Code hosting page for the details.
Also hosting a different coding agent? The same API runs Anthropic's Claude Code and any public Docker image. Full API details live on the Cloud page.