Gemini CLI,
in a real control center.
Run Google’s Gemini CLI in real terminals next to Claude Code, Codex, Kimi and Hermes — with a check on every shell command, a live map of the files it touches and its tokens counted per project.
Gemini and the others,
in one window.
The real Gemini CLI,
with a control center around it.
A real terminal
Every session is the real Gemini CLI in its own terminal, with everything you know from it.
Checked before it runs
Gemini starts with its own settings file containing OpenCognit’s check hook. Dangerous commands are refused, others wait for your OK. Your ~/.gemini stays untouched.
Visible in the Neurons
When Gemini reads or changes a file, a pulse runs to it in Gemini’s colour — live, next to the other agents.
Tokens counted
The Usage page reads Gemini’s local logs: tokens and API value per day, project and model.
Speak instead of type
Hold Ctrl+Space, speak, let go — your words land in the Gemini terminal. Recognised on your own computer.
Next to the others
Gemini, Claude, Codex, Kimi and Hermes in one grid. You see at a glance who works and who waits for you.
Three steps, no new account.
Install Gemini CLI
As usual, and log in with Google.
Install OpenCognit
Linux (AppImage, .deb) or Windows. It finds Gemini CLI by itself.
Open a project
Click “+”, pick Gemini CLI and give it a task.
Questions about Gemini CLI
Do I need a Gemini API key?
No. OpenCognit starts the Gemini CLI you already have installed, with your own login.
Does OpenCognit change my Gemini settings?
No. Sessions get their own settings file; your ~/.gemini stays as it is.
Can Gemini be the orchestrator?
No — Gemini runs as a terminal agent. Threads and the orchestrator are available with Claude Code, Kimi Code and Hermes.
Does it run on macOS?
Not yet. OpenCognit is available for Linux and Windows.
What does it cost?
Free with up to 4 agents and 2 projects. Pro is a one-time purchase — no subscription on top of your plan.
