Skip to main content
Give agents work, follow their progress, and steer them from the web app at app.opengeni.ai. You can also self-host. To add agents to your own product instead, see Embed with your coding agent.
1

Create an account

Sign up, choose Run agents in the cloud, and name your organization. You get a Personal workspace. Invite teammates from settings.
2

Choose a model

New verified accounts get $10 in trial credits, so you can start chatting right away. You can also connect:
  • a ChatGPT/Codex or SuperGrok subscription (for ChatGPT, first turn on device code login under Settings → Security);
  • a provider key such as OpenAI, Azure OpenAI, OpenRouter, or Vercel AI Gateway; or
  • prepaid Opengeni credits.
You can skip this and do it later from settings. Admins choose which workspaces can use each model in organization Models settings.
3

Start a session

Describe the task. Optionally attach a repository, upload files, or pick tools. The session runs in a managed sandbox.
4

Follow and steer

Progress streams in as the agent works. Messages you send during a turn queue behind it; Steer delivers yours immediately. Pause holds the session and Cancel ends it.
5

Answer the agent

When a tool needs approval or the agent asks a question, the session waits for your answer, then continues.

​
Keep it going

Give a session a goal with success criteria, and the agent keeps working until it completes the goal or pauses with a reason. See Goals.

​
Use your own computer

Connect a laptop or server and run sessions on it, with your files and git credentials. See Connect a machine.

​
Call it from code

Everything in the web app is available through the API. See Embed manually and the SDK reference.