Jules Review 2026: Google's Async AI Coding Agent

What Is Jules?

Jules is Google’s asynchronous autonomous coding agent. It launched in preview at Google I/O 2025 and exited beta in August 2025. You describe a task, Jules plans and executes it in an isolated Google Cloud VM, and you come back later to review the completed work as a GitHub pull request. It is powered by Gemini models.

What Sets It Apart

Jules is built around a fundamentally different workflow than most AI coding tools. Instead of requiring you to watch while AI writes code, Jules is async-first: describe the work, move on to something else, and return to a completed PR. For developers juggling multiple responsibilities, this reclaims the time other tools demand you spend waiting.

GitHub integration is deep and practical. Jules reads issues for context, clones your repo, installs dependencies, makes changes, runs your test suite for verification, and opens a pull request with a summary of what changed and why. Everything fits naturally into existing team workflows that already use GitHub for code review. You can trigger tasks from issue comments, the web interface, or programmatically via the Jules API.

The CLI and API are what turn Jules from a developer tool into development infrastructure. Trigger tasks from CI/CD pipelines when builds fail, create tasks from monitoring alerts, or batch-process tasks from a script. The file selector lets you scope tasks to specific directories, and the memory feature persists project conventions across sessions.

The limitations are equally clear. Jules is GitHub-only — no GitLab, Bitbucket, or self-hosted Git support. It runs exclusively on Gemini models with no BYOK option. There is no local IDE integration; all interaction happens through the web interface, CLI, or API. And task-based daily limits can be restrictive during intensive development sessions.

Who It’s For

• Developers who want to delegate tasks and review results later rather than watch AI work in real time
• GitHub-centric teams that want AI coding integrated into their existing PR workflow
• DevOps and platform engineers looking to automate coding tasks via API
• Not suited for teams on GitLab or Bitbucket, or developers who want real-time inline AI assistance in their editor

Sources

• Jules by Google
• Jules Documentation
• Jules Pricing