Free Introductory
$0/mo
15 tasks/day, 3 concurrent
Jules is Google's asynchronous AI coding agent. Assign a GitHub Issue (or describe a task in plain language) and it independently clones the repo into an isolated Google Cloud VM, plans the work, writes a fix, runs tests, and opens a pull request — all without the developer staying online. GA since August 2025, powered by Gemini 3 Pro on paid tiers and Gemini 3 Flash on free. Bundled with Google AI Pro and Ultra subscriptions, plus a Jules Tools CLI for terminal workflows.
+Pros
- Fire-and-forget async workflow: assign a GitHub Issue or task and Jules independently clones, plans, codes, tests, and opens a PR — no need to stay online or monitor progress.
- Isolated Google Cloud VMs per task ensure clean, reproducible environments with no side effects between runs, and Environment Snapshots allow reusing common setups across tasks.
- CI self-correction loop: if the pipeline fails on a newly opened PR, Jules detects the error, applies a fix, and re-pushes without human intervention — reducing manual review cycles.
- Generous free tier (15 tasks/day, 3 concurrent, powered by Gemini 2.5 Pro) makes it accessible for individual developers and small teams to try without upfront cost.
- Dual-model architecture with Gemini 3 Pro on paid tiers delivers strong reasoning for planning and complex multi-file changes, while the Critic Agent provides automated self-review.
−Cons
- GitHub-only as of 2026 — no support for GitLab, Bitbucket, or other code hosting platforms, which limits its usefulness for teams on alternative platforms.
- Vendor lock-in to Google Cloud infrastructure and Google AI subscriptions — both the Pro ($19.99/mo) and Ultra ($249.99/mo) plans are tied to Google AI subscriptions.
- Limited to well-defined, testable tasks with clear requirements — Jules excels at bug fixes and version bumps but struggles with ambiguous, architectural, or highly creative work.
Pricing
Google AI Pro
$19.99/mo
100 tasks/day, 10 concurrent
Google AI Ultra
$249.99/mo
Unlimited tasks, 60 concurrent
Jules is Google's asynchronous AI coding agent, originally launched in Google Labs in December 2024, opened to public beta at Google I/O in May 2025, and made generally available worldwide in August 2025. Its core design principle is fire-and-forget: assign a GitHub issue or describe a task, and the agent independently handles the entire resolution cycle — cloning the repository into an isolated Google Cloud VM, planning the work, writing the code, running the test suite, and opening a pull request for human review. During its public beta phase, Jules generated more than 140,000 publicly shared code contributions.
Jules runs on a dual-model configuration: Gemini 3 Pro (formerly Gemini 2.5 Pro) for planning and complex reasoning on paid tiers, with lighter Gemini 3 Flash work on the free tier. Each task runs in a fully isolated Google-managed VM with its own checkout of your repo — long-running work never ties up your laptop, and Environment Snapshots let common setups be reused across tasks. A CI feedback loop closes the cycle: if the pipeline fails on the newly opened PR, Jules picks up the error, applies a fix, and re-pushes the commit without human intervention. Private-repo code is never used to train Google's models.
Beyond the GitHub-native flow, Jules added several 2026 features: a Critic Agent that reviews its own plan and output, Interactive Plan Mode for editing the agent's plan mid-execution, automated web app testing with visual screenshots, multimodal support for visual output, audio changelog summaries (listenable commit history), and the Jules Tools CLI for terminal-based task creation and monitoring. Pricing is tightly coupled with Google AI subscriptions: a free Introductory tier, Google AI Pro for daily individual use, and Google AI Ultra for high-throughput parallel agent workloads.
Key Features
• Async GitHub Issue → PR: fire-and-forget, no supervision required
• Isolated Google Cloud VMs per task — clean, reproducible, no side effects
• Dual-model: Gemini 3 Pro on paid tiers, Gemini 3 Flash on free
• CI feedback loop: Jules picks up pipeline failures and self-corrects the PR
• Interactive Plan Mode: edit the agent's plan before, during, and after execution
• Critic Agent: self-review of plan and changes
• Environment Snapshots: reuse environment configs across tasks
• Automated web app testing with visual screenshots
• Audio changelog: listenable summaries of recent commits
• Jules Tools CLI: create and monitor tasks from the terminal, pipe `gh` issues into Jules
• Multi-file editing, full test execution, automatic PR creation
• Privacy: private-repo code is not used for model training
• Free Introductory: 15 tasks/day, 3 concurrent
• Google AI Pro: 100 tasks/day, 15 concurrent
• Google AI Ultra: 300 tasks/day, 60 concurrent
• GitHub-only (no GitLab/Bitbucket support as of 2026)
Version History
Added Jules Tools CLI for terminal-based task creation and monitoring (pipe gh issues into Jules), audio changelog for listenable summaries, Gemini 3 Pro upgrade on paid tiers, increased concurrency limits.
Introduced Critic Agent for automated self-review, Interactive Plan Mode for editing plans mid-execution, automated web app testing with visual screenshots, multimodal support.
General availability worldwide — full async GitHub Issue-to-PR workflow, isolated Cloud VMs, dual Gemini 2.5 Pro/Flash, CI self-correction loop, Environment Snapshots.
Signature Snippet
A developer assigns GitHub Issue #847 ('Login button unresponsive on Safari mobile') to Jules. Jules clones the repo into an isolated VM, reproduces the bug, traces it to a CSS pointer-events conflict, writes the fix across two files, runs the existing test suite (all green), adds a regression test, and opens a pull request with a detailed description — all within 8 minutes, while the developer was in a meeting.More in this Space
Prime Agent (Prime Intellect)
Prime Agent is an open-source (MIT) self-improving coding and research harness by Prime Intellect, built on a Recursive Language Model that runs inside a persistent IPython REPL and a Continual Harness that refines its own state via /refine. It took the same frontier weights from 30.2% to 95.5% on ARC-AGI-3 RHAE Best@1, powered daemon-backed sessions, and ships no hosted offering.
OpenCodeReview (ocr)
OpenCodeReview (ocr) is Alibaba's open-source AI code-review CLI. It reads git diffs and drives an LLM agent through a hybrid architecture where deterministic pipelines handle file selection, rule matching, and comment positioning while the agent does the reasoning. Built on two years of internal use at Alibaba scale, it ships a multi-language ruleset and consumes roughly 1/9 of the tokens of general-purpose agents.
Delta (by Zed)
Multiplayer environment from Zed for coding with agents. Threads unify the agent conversation, the code edits, and human review in one shareable artifact, replacing the branch/PR loop. Built on DeltaDB, a CRDT-based version control layer that records every edit between commits alongside the prompts and reasoning that produced them.