Sweep is an AI coding assistant built specifically for JetBrains IDEs, including IntelliJ IDEA, PyCharm, Android Studio, WebStorm, PhpStorm, and Rider. It combines a next-edit autocomplete model with an in-editor AI agent that can carry out multi-file coding tasks directly inside the IDE.
Sweep AI, Inc. originally launched Sweep through Y Combinator’s Summer 2023 batch as a GitHub bot that converted written issues into pull requests automatically. The company has since shifted its product toward a JetBrains-first model, and the original GitHub issue-to-PR workflow is no longer the core offering.
The current product centers on two features. Tab autocomplete predicts a developer’s next edit rather than just completing the current line, using a custom model built by the company. The Agent handles broader tasks, such as multi-file changes and reviewing diffs between branches, using models Sweep says it built in-house.
Sweep reports a 4.9-star rating and more than 40,000 installs on the JetBrains Marketplace. It targets individual developers and engineering teams who work primarily in JetBrains tools and want AI assistance without switching to a different editor.
Pricing
Sweep offers a free trial with 1,000 autocompletes and $5 in included API credits. Paid plans start at Basic for $10 per month, which includes unlimited autocomplete, followed by Pro at $20 per month with a larger API credit allowance and priority support, and Ultra at $60 per month for the highest credit allocation. API credits are consumed separately for agent features like chat and code generation, and accounts can be configured to top up automatically. Sweep also offers discounted pricing for students, educators, and open source maintainers on request.
* Disclaimer: Please note that pricing information may not be up to date. For the most accurate and current pricing details, refer to the official website.
Key Features
-
✓
Next-edit autocomplete that predicts upcoming code changes, not just line completions
-
✓
In-editor AI agent for multi-file coding tasks within JetBrains IDEs
-
✓
Codebase indexing for context-aware suggestions across a project
-
✓
AI code review for comparing diffs between branches
-
✓
Privacy Mode with zero data retention and no training on user code
-
✓
Remote MCP server support with OAuth 2.0 and 2.1 authentication
Use Cases
Faster Everyday Autocomplete
Developers writing routine code in PyCharm or IntelliJ rely on Tab suggestions to complete multi-line edits instead of typing them manually. This speeds up repetitive coding patterns like boilerplate and small refactors.
In-IDE Multi-File Changes
A developer describes a task to the Agent, such as updating a function signature across several files. The agent applies the edits directly in the IDE for review before committing.
Reviewing Branch Diffs
Before opening a pull request, a developer uses Sweep’s AI code review feature to compare changes between branches. This surfaces potential issues without leaving the IDE.
Working on Regulated Codebases
Teams handling sensitive or proprietary code enable Privacy Mode so that code is never stored or used to train models. This lets them use AI suggestions while limiting data exposure.
Connecting External Tools to the Agent
A team configures a remote MCP server so the agent can pull in context from other internal tools during a coding session. OAuth 2.0/2.1 support governs access to these connections.
Strengths & Weaknesses
Strengths
Holds a 4.9-star rating with more than 40,000 installs on the JetBrains Marketplace.
Covers the full range of JetBrains IDEs, including IntelliJ, PyCharm, WebStorm, and Rider.
Offers SOC 2 compliance and a Privacy Mode with zero data retention.
Uses a purpose-built Tab model designed specifically for fast next-edit predictions.
Supports remote MCP servers with modern OAuth authentication for extending agent capabilities.
Weaknesses
Only works within JetBrains IDEs, so developers on VS Code or other editors cannot use it.
Unlimited autocomplete requires a paid plan starting at $10 per month.
Agent and chat features draw on metered API credits, which can add cost for heavy users.
The original GitHub issue-to-pull-request workflow has been discontinued, limiting options for teams that relied on it.
Who Is This For?
JetBrains Developers: Engineers who already work in IntelliJ, PyCharm, or similar IDEs and want faster autocomplete without switching editors.
Engineering Teams with Security Requirements: Teams that need SOC 2 compliance and zero data retention for AI tooling used on proprietary code.
Individual Developers Doing Refactors: Programmers who regularly perform multi-file edits and want an agent to apply changes directly in the IDE.
Teams Building on MCP: Organizations that want to connect external tools and data sources to their coding agent through remote MCP servers.
Frequently Asked Questions
Which IDEs does Sweep support?
Sweep supports all JetBrains IDEs, including IntelliJ IDEA, PyCharm, Android Studio, WebStorm, PhpStorm, Rider, CLion, and RubyMine.
Does Sweep still work as a GitHub bot that opens pull requests from issues?
No. That workflow was part of Sweep’s original product but has since been deprecated in favor of the current JetBrains autocomplete and agent plugin.
Does Sweep train AI models on my code?
Sweep offers a Privacy Mode that prevents your code from being stored or used for model training. It can be enabled in account settings.
What does the free trial include?
The free trial includes 1,000 autocompletes and $5 in API credits, with all JetBrains IDEs and Privacy Mode available.
How much does Sweep cost after the trial?
Paid plans are Basic at $10 per month, Pro at $20 per month, and Ultra at $60 per month, each with increasing API credit allowances.
How do API credits work?
API credits are consumed by agent features like chat and code generation, while autocomplete is unlimited on paid plans. Usage can be tracked in the account balance tab.
Can I cancel my subscription at any time?
Yes. Cancellations take effect at the end of the current billing period, with no long-term commitment required.
Does Sweep offer discounts for students or open source maintainers?
Yes. Sweep offers special pricing for students, educators, and open source maintainers who contact the company directly for verification.
Is Sweep SOC 2 compliant?
Sweep states it is SOC 2 compliant and positions itself as enterprise-ready, with details available in its security documentation.
Can Sweep’s agent connect to external tools?
Sweep added support for remote MCP servers with OAuth 2.0 and 2.1 authentication, allowing the agent to connect to external tools and data sources.
Sweep integrates directly into JetBrains IDEs as a plugin, covering IntelliJ IDEA, PyCharm, Android Studio, WebStorm, PhpStorm, Rider, and other JetBrains products. It also supports remote MCP servers with OAuth 2.0 and 2.1 authentication, allowing the agent to connect to external tools. No GitHub Actions or other CI/CD pipeline integrations are currently listed on Sweep’s public site.