SWE-agent
A research-born, open-source autonomous agent that takes a GitHub issue and tries to fix it, using any model through an agent-computer interface.
- Free plan
- Autonomous agent
- Mixed sources
- Verified 13 Nov 2026Sources are dated on this profile
- 3 sourcesLinked, typed and dated
- Vendor and third-party sourcesHow this profile was compiled
- Not an endorsementStack18 has not tested this product
What is SWE-agent?
SWE-agent is an open-source (MIT) autonomous agent from academic researchers that takes a GitHub issue and attempts to resolve it, using a purpose-built agent-computer interface and any language model. It grew out of SWE-bench research and is actively maintained.
- Best forResearchers and benchmarkers
- PricingFree and open source; you pay only for model usage. Free plan available.
- Main limitationResearch-oriented rather than a polished product.
- AlternativesOpenHands · Aider · Claude Code · compare all
SWE-agent overview: who it is for and what it does.
SWE-agent gives a language model a set of tools designed for software tasks, such as viewing, searching and editing files, and lets it work autonomously on an issue in a sandboxed environment. It is widely used for research and benchmarking on SWE-bench.
It is free and open source with about 19,000 GitHub stars and recent commits; you pay only for model usage.
Main problem it solves
Automates fixing well-scoped GitHub issues and supports research on coding agents.
Typical customer segment
Researchers, platform teams and developers.
Target users
- Terminal-first developers
- Open-source contributors
- Developers who bring their own model keys
Primary use cases
- Automatic GitHub issue fixing
- Benchmarking agents on SWE-bench
- Research on agent interfaces
- Batch issue runs
What makes it distinct
- Research-backed agent-computer interface.
- Autonomous issue resolution.
- Actively maintained open source.
SWE-agent features explained.
Agent
- Agent-computer interface
- Tools designed for LLM software work.
- Issue resolution
- Works autonomously on GitHub issues.
Models
- Any model
- Multi-model support.
SWE-agent pricing and plans.
Figures reported consistently by independent sources; vendor page not reviewed directly. Checked 13 Nov 2026 against SWE-agent (GitHub) · OpenAlternative · Toolhalla.
Pricing model
Free open source; pay model usage.
Free plan
Yes. Free and open source.
Free trial
Not applicable.
SWE-agent plans and prices (USD)
-
Open source
Free
n/a
MIT licence; pay for models.
No paid plans; costs are model API usage.
Prices change. Confirm current plans and terms on the vendor's site before you decide.
Visit official website (external) ↗See alternativesSWE-agent integrations, API and platforms.
Important integrations
- GitHub
- Docker sandbox
- Any LLM provider
Supported platforms
- Terminal (Linux, macOS)
- Docker
API
Open source; configured with provider API keys.
Data import and export
Not reviewed for this profile.
Integration availability can depend on plan. Confirm with the vendor before you rely on a specific connector.
How SWE-agent works as an AI product.
Attributes are listed only where documented. We do not infer autonomy levels, certifications or model capabilities.
Classification: Autonomous agent. Completes multi-step tasks with tools and limited supervision, as documented by the vendor. Developer: SWE-agent project (Princeton and Stanford researchers).
| Attribute | Details |
|---|---|
| Supported models | Multi-model support. [3] |
| Tools and integrations | Not documented in the sources reviewed. |
| Ability to execute actions | Autonomously edits code to resolve GitHub issues. [3] |
| Execution environment | Sandboxed environment. [3] |
| Human-in-the-loop controls | Not documented in the sources reviewed. |
| Permissions and access | Not documented in the sources reviewed. |
| Memory and context | Not documented in the sources reviewed. |
| Deployment options | Not documented in the sources reviewed. |
| Data handling and privacy | Not documented in the sources reviewed. |
| Security and compliance | Not documented in the sources reviewed. |
| Pricing and usage limits | Not documented in the sources reviewed. |
| Languages and platforms | Not documented in the sources reviewed. |
| Limitations and known constraints | Not documented in the sources reviewed. |
Primary tasks
- Terminal coding tasks
- Multi-file edits
- Running commands and tests
Best-fit workflows
- Terminal-first developers
- Open-source and BYOK setups
- Scripting and automation
SWE-agent pros and cons.
Strengths, limitations and best fit. Each point cites a source; editorial summaries are not customer reviews.
Strengths
Limitations
Best for
- Researchers and benchmarkers
- Platform teams automating issue fixes
- Developers exploring autonomous agents
May be less suited to
- Teams wanting a polished IDE experience
- Non-technical users
Top SWE-agent alternatives and competitors.
SWE-agent alternatives
Developers look for SWE-agent alternatives when they want a polished product with IDE integration, team…
Explore →Related productOpenHands
An open-source platform for AI software development agents, with an SDK, CLI, local GUI, hosted cloud and a…
Explore →Related productAider
An open-source terminal tool for AI pair programming that edits your local git repository and commits each…
Explore →Related productClaude Code
Anthropic's agentic coding tool for the terminal, IDEs, a desktop app and the web, built around…
Explore →SWE-agent compared with similar AI coding agents.
Summaries come from each profile. This is a side-by-side aid, not a ranking.
| Product | Interface | Model choice | Free tier | Pricing model | Human oversight |
|---|---|---|---|---|---|
| SWE-agent | Terminal and Docker sandbox | Any model | Yes (open source) | Free software; pay model usage | Autonomous; review output before merging |
| Aider | Terminal (command line) | Almost any LLM, including local models | Yes (open source) | Free software; pay model usage | Git commits can be reviewed and undone (third-party) |
| Amp | CLI and editor extensions | Several frontier models | Yes (ad-supported daily grant) | Credits and subscriptions | Not documented in sources reviewed |
| Claude Code | Terminal, VS Code, JetBrains, desktop app, web | Claude models (Anthropic) | Not documented in sources reviewed | Included in paid Claude plans, or usage-based API billing | Permissions, sandboxing and checkpoints documented |
| Codebuff | Terminal (command line) | Not documented in sources reviewed | Yes (500 credits a month) | Free credits; paid $100 to $500 | Not documented in sources reviewed |
| Crush | Terminal TUI | Many providers, switchable mid-session | Yes (free software) | Free software; pay model usage | Not documented in sources reviewed |
SWE-agent vs competitors: head-to-head comparisons.
SWE-agent discussions and buying questions.
SWE-agent questions and answers
What is SWE-agent used for?
SWE-agent is an open-source (MIT) autonomous agent from academic researchers that takes a GitHub issue and attempts to resolve it, using a purpose-built agent-computer interface and any language model. It grew out of SWE-bench research and is actively maintained.
Is SWE-agent free?
Yes. Free and open source.
How much does SWE-agent cost?
Free and open source; you pay only for model usage.
Which models does SWE-agent support?
Multi-model support.
What are the limitations of SWE-agent?
Research-oriented rather than a polished product. No team or admin features. Model costs are separate.
Which platforms does SWE-agent run on?
SWE-agent is documented as available on: Terminal (Linux, macOS), Docker.
Is SWE-agent an AI agent or an AI assistant?
We classify SWE-agent as: Autonomous agent. Completes multi-step tasks with tools and limited supervision, as documented by the vendor. The classification is based on vendor documentation and is not an endorsement.
What are the main alternatives to SWE-agent?
The main alternatives to SWE-agent covered on Stack18 are OpenHands, Aider, Claude Code and Devin. See the SWE-agent alternatives guide for a side-by-side table.
When was the SWE-agent profile last verified?
Sources for this profile were last reviewed on 13 Nov 2026. 3 sources are listed under Sources, and prices and plans can change, so confirm with the vendor.
Where this information comes from.
- SWE-agent repository SWE-agent (GitHub) Official Accessed 13 Nov 2026
- Plandex vs SWE-agent OpenAlternative Third-party Accessed 13 Nov 2026
- Plandex vs SWE-Agent Toolhalla Third-party Accessed 13 Nov 2026
Further official pages: SWE-agent pricing (external)
Plans, features and status were checked against SWE-agent's published pages and the third-party sources listed. Agent behaviour was not tested, and security certifications were not independently verified.
Stack18 has not tested this product or received anything from the vendor for this profile. Editorial summaries of vendor information are not endorsements. Confirm details with the vendor before buying.
Evaluating SWE-agent?
Check the vendor's current pricing and terms, compare it with similar products, or read how others approach the decision. If off-the-shelf is not the answer, Stack18 builds around how your business works.
Logos provided by Logo.dev. Product names and logos are trademarks of their respective owners and appear for identification only; they do not imply endorsement.