Google Jules is an autonomous AI coding agent that connects to GitHub to handle coding tasks asynchronously.
Google Jules is an autonomous coding agent that works inside GitHub repositories. Connect a repository and branch, then describe a task in a detailed prompt or label a GitHub issue with "jules." The agent fetches the code, clones it to a Cloud VM, and drafts a plan using Google's Gemini models. It then provides a diff of proposed edits for review. After approval, Jules opens a pull request that can be merged and published.
The tool is built for developers who want to offload routine coding work and stay focused on larger tasks. It runs asynchronously, so you can assign tasks and return later. Plans differ by daily task limits, concurrency, and model access. A free tier offers 15 tasks per day and 3 concurrent tasks with Gemini 2.5 Pro; paid plans increase limits and unlock higher access to newer models like Gemini 3 Pro.
Jules integrates with GitHub. You select a repository and branch, and Jules fetches and clones the code to a Cloud VM to work on it.
Yes. You can use the "jules" label on a GitHub issue to assign that task to Jules.
Jules uses Gemini models. The free tier uses Gemini 2.5 Pro, while paid plans offer access to newer models like Gemini 3 Pro.
Jules provides a diff of the proposed edits. You can browse and approve the changes before Jules creates a pull request.
No. Jules creates a pull request, but you must approve and merge it yourself.
The free tier includes 15 tasks per day and 3 concurrent tasks. Paid plans offer 100 tasks per day with 15 concurrent tasks, or 300 tasks per day with 60 concurrent tasks.
Yes, there is a free tier with limited daily tasks and concurrency, using Gemini 2.5 Pro.
Yes, concurrent task limits depend on your plan: 3 on the free tier, 15 on the mid-tier, and 60 on the highest tier.
