Jules
Google's asynchronous coding agent that turns prompts into reviewed GitHub pull requests
Editorial take
Jules is the easiest way to experience autonomous coding agents today: the free tier is real, the PR workflow is disciplined, and the CI auto-fix is something even pricier rivals lack. Treat it as a tireless junior engineer for well-scoped tasks and it earns its keep; hand it ambiguous feature work and you will spend the savings on review.
What is Jules?
Jules is Google's autonomous coding agent, and it works differently from an editor assistant: you point it at a GitHub repository, describe the task, and walk away. Jules clones the repo into an isolated Google Cloud VM, writes a plan for your approval, makes the changes, runs a built-in Critic review pass, and opens a pull request. Since early 2026 it also fixes its own failing CI runs on those pull requests and resubmits.
It graduated from beta in August 2025 and has grown a full toolchain: the Jules Tools CLI (npm), an alpha REST API that lets you trigger tasks from Slack or CI/CD, MCP support, scheduled recurring tasks, per-repo memory of your preferences, and an AGENTS.md convention for repo-specific context. Labeling a GitHub issue with "jules" is enough to hand it the work. The free tier runs on Gemini 3 Flash; paid Google AI plans unlock Gemini 3 Pro class models.
The honest positioning: Jules is for well-scoped, fire-and-forget work such as bug fixes, tests, refactors and dependency bumps, executed in parallel while you do something else. It is not an interactive pair-programmer, it only speaks GitHub, and its VM cannot keep long-running processes like dev servers alive. Reviewers consistently note that its output must be checked like a junior engineer's.
Best for
Teams offloading bug fixes, tests and dependency bumps as background pull requests
Key strength
Fire-and-forget GitHub pull requests with a built-in critic and CI auto-fix
Score breakdown (out of 5)
What you would use it for
- →Clearing a backlog of small bug fixes as parallel pull requests
- →Writing missing tests across a repository
- →Dependency and framework version bumps with CI verification
- →Scheduled recurring maintenance like lint cleanups
Pros & Cons
👍 Pros
- ✓Genuinely useful free tier: 15 tasks a day with the full PR workflow
- ✓Parallel background execution frees you while it works
- ✓CI auto-fix closes the loop most async agents leave open
- ✓Included in Google AI plans rather than a separate subscription
👎 Cons
- ✗GitHub is the only supported code host
- ✗Not an interactive assistant: you steer between tasks, not during them
- ✗VM cannot run long-lived processes such as dev servers
- ✗Independent tests report roughly two thirds of tasks merge-ready on first attempt, so review remains mandatory
Key Features
- ✓ Async workflow: clone to cloud VM, plan, diff, pull request
- ✓ Built-in Critic agent reviews changes before you see them
- ✓ Auto-fixes failing CI on its own pull requests
- ✓ Jules Tools CLI and alpha REST API for Slack and CI/CD triggers
- ✓ Scheduled recurring tasks and proactive task suggestions
- ✓ Per-repo memory plus AGENTS.md context convention
- ✓ Assign work by labeling a GitHub issue with jules
Available on
Integrates with
Jules Pricing
✅ Jules has a free plan - no credit card required to start.
Free
- ✓15 tasks per day, 3 concurrent
- ✓Gemini 3 Flash base model
- ✓Full plan-review-PR workflow
- ✓CLI and GitHub integration
Jules in Pro (Google AI Pro)
- ✓100 tasks per day, 15 concurrent
- ✓Gemini 3 Pro class models
- ✓Bundled with the full Google AI Pro plan
Jules in Ultra (Google AI Ultra)
- ✓300 tasks per day, 60 concurrent
- ✓Priority access to the newest models
- ✓Bundled with Google AI Ultra
Jules vs Competitors
Developer resources
Related Tools
Claude is the best AI assistant for long-document work and research where accuracy matters. Its 200K context window and resistance to sycophancy make it the honest choice when you need a thinking partner rather than an agreeable assistant. Start on the free tier; the Pro plan at $20/month is worth it if you regularly hit daily limits.
ChatGPT is the most capable all-in-one AI assistant available. Its combination of image generation, voice mode, code execution, web browsing, and a large custom GPT ecosystem makes it the default choice for users who want one tool that handles everything. The tradeoff is that it is rarely the deepest tool on any single task.
Cursor is the most powerful AI code editor available. Its multi-file editing is genuinely ahead of plugin-based alternatives, making large-scale refactors practical in ways GitHub Copilot cannot match. At $20/month for Pro it is a meaningful upgrade for any developer doing more than small isolated changes.
GitHub Copilot is the right starting point for most developers adopting AI-assisted coding. It integrates into whatever editor you already use, requires no workflow change, and has a usable free tier. The Pro plan at $10/month is the lowest-cost path to unlimited AI completions, making it the practical default for individuals.
This page contains affiliate links. We may earn a commission at no extra cost to you. Learn more.