How to use your Google AI Pro subscription inside Claude Code (and Codex, and the Gemini CLI)
If you're like me, you pay for one AI subscription and do most of your coding in another tool. I use Claude Code all day, and I also have Google AI Pro. Gemini 4 Argon, Deep Research, Flow, NotebookLM and Jules were all included, and I barely touched them, because switching tabs breaks the flow of work. So I wrote an open-source MCP server, gemini-pro-mcp , that lets Claude (or Codex, or the…
For AI enthusiasts who use a single subscription for coding, Google AI Pro presents an efficient solution. Gemini models like Argon, Deep Research, Flow, NotebookLM and Jules are included in the plan, but switching tabs interrupts the coding flow. To address this issue, the open-source MCP server gemini-pro-mcp was created, enabling Claude (or Codex, or the Gemini CLI) to directly call the plan.
MCP, or Model Context Protocol, is a standard method for granting AI agents tools. An MCP server offers functions such as gemini_ask(prompt), which a client, like Claude Code, can utilize. The server is installed once and can be used by any MCP client. Upon setup, Gemini models are accessible without additional API charges. With gemini_ask, gemini_research and gemini_code, Gemini can be utilized for tasks like reviewing code for race conditions, conducting research with sources, or editing files in a designated folder with a sandboxed terminal.
Gemini Notebooks (formerly NotebookLM) allows creating a notebook, adding PDFs or URLs, asking questions based on those sources, and generating Audio Overviews, mind maps, quizzes, or slide decks. Jules enables starting a session on a GitHub repo, approving its plan, and sending messages through the official API. Flow videos, though experimental, convert a 2-minute video into a shot list, generates each clip in Google Flow with monthly credits, and combines them into one file.
The server is MIT-licensed and available on PyPI and the official MCP Registry. Contributions are welcome for adding new tools.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
