Wow. This just got even more insane.
Google has officially taken Jules out of beta and unleashed an absolute coding beast — something many people are calling Jules 2.0.
The new Jules is a huge evolution in what agentic development can look like — combining planning, sophisticated autonomy, and serious engineering maturity in one system.
Google is clearly getting dead serious about developer tooling. No more wishy-washy experiments. No more half-steps. Jules is here to stay.
Your AI teammate, smarter than ever
Jules is still that genius AI coding agent that can understand your intent, plan steps, and execute complex coding tasks — all asynchronously, like a super-smart teammate working in the background.
But with this update Jules is faster, more capable, and more reliable.
It uses Gemini 2.5 Pro for reasoning and planning — serious brainpower for tough coding problems. And with Gemini 2.5 Flash on the horizon, even crazier speeds are coming.
The brilliant features in Jules 2.0
Critic-augmented generation
Before showing you results, Jules now reviews its own work with a built-in critic.
If the critic finds issues, Jules fixes them before you even see the code. Think of it as a safety net built right in.
GitHub issues integration
You can assign Jules tasks straight from GitHub Issues. It turns tickets into working pull requests — automatically.
Reusable setups
Reruns are faster than ever because Jules remembers past contexts, so repeat workflows don’t have to start from scratch.
Multimodal support
Not just code anymore — Jules can handle a wider range of inputs and contexts.
Audio changelogs
Jules can tell you what changed, so you can keep up just by listening.
How Jules works
When you give Jules a task, it clones your repo into a secure Google Cloud VM, isolated from your live code. There, it experiments safely with full context of your project.
It then:
- Plans the steps.
- Explains its reasoning.
- Generates a “diff” of changes.
- Opens a pull request on GitHub for you.
You stay in full control — reviewing, approving, or modifying its work before merging.
Jules shines at all the boring but critical coding chores:
- Writing tests
- Fixing bugs
- Adding features
- Bumping dependencies
- Updating docs
Availability & plans
Jules is now generally available worldwide. No waitlist. No invite-only beta.
- Free Plan: 15 tasks/day, 3 concurrent runs.
- Pro Plan: 100 tasks/day, 15 concurrent runs.
- Ultra Plan: 300 tasks/day, 60 concurrent runs.
This scaling makes it usable for everyone from hobbyists to full-on teams.
The bigger picture
This isn’t about replacing developers — at least for now.
Google is framing Jules as a force multiplier: freeing you from grunt work so you can focus on architecture, creativity, and real problem-solving.
Jules is less about taking jobs, more about changing the workflow: you offload the boring, repetitive stuff to an AI agent that works in the background, while you design the future.
Final thoughts
Jules was already mind-blowing in beta.
But with these new updates and the general release, Google just moved the conversation forward to push the boundaries of agentic development.
If you’re a developer this isn’t optional anymore. It’s the start of a new way of building.
Check it out, experiment, and get ready — because coding just changed again.