Gemini CLI GUI · free

A free desktop app for Gemini CLI.

Google’s open-source coding agent for the terminal. Tesseract gives it a window on Mac, Windows and Linux — beside a code editor, a terminal and a live preview, with a plan you approve before it acts.

Updated · Made by Google · Gemini CLI’s own site

The model menu in a conversation, with recent models and models grouped by maker.
Set up

Gemini CLI running in three steps.

  1. Download Tesseract. Free for Mac, Windows and Linux; it installs and updates itself.
  2. Pick Gemini CLI. On the first-run screen or in Settings. Tesseract installs it for you, or uses the copy already on your computer.
  3. Sign in. Use a Gemini API key. Google sign-in now works only with a Gemini Code Assist licence: Google stopped serving personal and Google AI Pro/Ultra accounts in Gemini CLI on June 18, 2026.
What works

Gemini CLI in Tesseract, feature by feature.

Read from the app’s own framework catalog — what Tesseract does with Gemini CLI today, not what any agent promises.

  • Asks you before a risky tool call
  • Pick the model for each message
  • Choose how hard it thinks
  • Send it images and files
  • Drives a real browser
  • Tesseract’s tools and your MCP servers
  • Skills and extensions
  • Subagents
  • Pick up a past conversation
  • Fork a conversation
  • Restore files to an earlier message
  • A reviewable diff of everything it changed
  • Shows how full its context window is
  • Compacts a long conversation to keep going
  • Keeps working when you step away
  • Scheduled, repeating jobs
  • Checks its own work before it finishes
  • Runs inside the operating system’s sandbox
  • The cost of every conversation
  • An integrated terminal
  • Search everything it has said
Around it

Everything an agent needs, in the same window.

  • A code workspaceThe Monaco editor with a file tree and tabs, an integrated terminal, and a live preview while it works.
  • A plan firstChat, Plan or Full access for each conversation, and an editable plan you approve before it acts.
  • A real browserA browser the agent can drive for forms and live pages, and your own signed-in browser through the extension.
  • Your other computersRun Gemini CLI on a server or another machine with tesseract-agent, and follow it from here.
  • Tools and skillsAny MCP server over stdio, HTTP or SSE, skills from ClawHub, and extensions with their own tools.
  • Free to useTesseract costs nothing. Your Google account or key works as it does anywhere else.
Questions

About Gemini CLI in Tesseract.

Is there a Gemini CLI GUI for Mac, Windows and Linux?
Yes. Tesseract is a free desktop app for all three that runs Gemini CLI in a window, with a code editor, a terminal, a live preview and a browser beside the conversation.
Does it cost anything to use Gemini CLI in Tesseract?
Tesseract is free. Use a Gemini API key. Google sign-in now works only with a Gemini Code Assist licence: Google stopped serving personal and Google AI Pro/Ultra accounts in Gemini CLI on June 18, 2026. Your plan or key with Google works as it does anywhere else, with its own limits; Tesseract adds no charge on top.
Can I use Gemini CLI and other agents at the same time?
Yes. Each conversation picks its own framework, so Gemini CLI can work in one conversation while Claude Code, Codex or any of the 13 others work in another.
Does Tesseract change how Gemini CLI works?
It runs Gemini CLI itself, installed on your computer, and adds what is around it: the window, the editor and terminal, a plan you approve, permission prompts and a record of the conversation you can search and pick up later.