Run Any Coding Agent and Model in the Cloud
Run Claude Code, Codex, Cursor, OpenCode and more in the cloud with Tembo, on the model you choose, using your own subscription or API keys.
Tembo runs the coding agents your team already uses in the cloud: Claude Code, Codex, Cursor, OpenCode, Pi, Amp, Grok Build and Antigravity. You pick the agent and the model for each session or automation, and you can pay for model usage with Tembo, with your own API keys, or with the Claude, ChatGPT or SuperGrok subscription you already have.
Every engineer has a favorite agent, and the best model changes every few months. Your platform shouldn't lock you into either choice.
Which coding agents can I run?
Tembo supports these coding agents, also called harnesses:
- Claude Code
- Codex
- Cursor
- OpenCode
- Pi
- Amp
- Grok Build
- Antigravity
Each one runs in its own sandbox: a dedicated Linux VM with your repositories, a desktop, a terminal and the tools your project needs. When the work is done, Tembo opens a pull request for your team to review. If you want to look around while the agent works, you can open the session's desktop in Tembo Computer.
Which models can they use?
Tembo's model catalog covers the major providers and the strongest open models. A few highlights from the current list:
| Provider | Models include |
|---|---|
| Anthropic | Claude Opus 5.5, Claude Fable 5.1, Claude Sonnet 5.5, Claude Haiku 5.5 |
| OpenAI | GPT-6.1 Sol, GPT-6 Astra, GPT-5.5 Pro, GPT-5.3 Codex |
| Gemini 3.1 Pro, Gemini 3.5 Flash | |
| xAI | Grok 4.7 |
| Cursor | Composer 2.5 |
| Open models | GLM 5.3, Kimi K3, MiniMax M3, DeepSeek V4 Pro |
Which models you see depends on the agent you pick and your workspace settings. The models docs have the full, current table.
Can I use my Claude or ChatGPT subscription?
Yes. Connect a subscription under Settings > Connected accounts, in the LLM subscriptions section:
- Claude (Connect Claude): use Anthropic models with Claude Code. This connection is personal, so it applies to your sessions only.
- ChatGPT (Connect Codex): use OpenAI models with Codex, OpenCode or Pi. A ChatGPT subscription can also be connected for the whole workspace, and a member's personal subscription takes priority over the workspace one.
- SuperGrok (Connect SuperGrok): use Grok models with OpenCode or Pi.
Subscriptions work during your free trial and on paid plans.
Can I bring my own API keys?
Yes. Add keys under Settings > Models > API Keys for OpenAI, Anthropic, xAI, Cursor, Amp or OpenRouter, or connect AWS Bedrock or GCP Vertex AI. Workspace admins choose per model whether it uses your key or Tembo managed inference, from the model's Inference dropdown.
BYOK inference costs $0 through Tembo. Sessions still use cloud VM compute, which is billed as usual.
How do I choose the agent and model?
There are three levels, and the most specific one wins:
- Workspace default. Admins set it in Settings > Models. It applies to new sessions and automations.
- Personal default. Each person can set their own in Settings > Preferences, so one engineer can default to Claude Code while another uses Codex.
- Per session or automation. Override the agent and model when you start a session, or in an automation's settings.
Admins can also turn individual models on or off for the organization in the Available Models list.
Why run coding agents in the cloud?
Running agents on a laptop ties them to that laptop. In the cloud, each session gets its own machine, so you can run several at once, close your laptop, and pick up the results later. Agents can also start from places other than your editor: a schedule, a Slack message, a Linear or Jira ticket, or a Sentry error, as covered in How to Keep Your Docs Up to Date Automatically.
FAQ
Can I run Claude Code in the cloud? Yes. Pick Claude Code as the agent in Tembo and it runs in a cloud sandbox with your repositories, on the Anthropic model you choose. You can use Tembo's managed models, your own Anthropic key, or your Claude subscription.
Can I use my ChatGPT subscription with Codex in Tembo? Yes. Connect it with Connect Codex under Settings > Connected accounts, and use supported OpenAI models with Codex, OpenCode or Pi.
Can different people on my team use different agents? Yes. Each person can set a personal default agent and model, and override it per session.
Does using my own API key cost anything in Tembo? BYOK inference costs $0 through Tembo. You pay your provider for the tokens, and Tembo bills the cloud VM compute the session uses.
Get started
Open Tembo, start a session, and pick your agent and model. For the full model list and setup steps, see the models docs.
Run any coding agent in the cloud
Tembo agents execute tasks in secure cloud environments and return reviewable output. Use any agent or model, run in parallel, and keep humans in control.
