Devin vs SWE-agent

Overview

Devin is an autonomous software engineer from Cognition. You give it a task in natural language — fix a bug, add a feature, migrate a dependency, write tests — and it works in its own cloud VM (or on your machine), browses documentation, runs commands, tests in its own browser and keeps going until the pull request is ready to merge.

Unlike autocomplete-style assistants, Devin is designed to own a task end to end: it plans, edits code, executes tests, reads error output and iterates. It is positioned as a teammate you delegate to rather than a tool you drive keystroke by keystroke.

Core capabilities

• End-to-end task execution — from reading the ticket to opening a PR, including running tests and fixing failures.
• Cloud or local execution — sessions run in isolated cloud VMs by default, or on your own machine when the work needs local resources.
• Built-in browser — Devin can open docs, dashboards and web apps to verify UI behavior and gather context that is not in the repo.
• Parallel sessions — run several Devins at once on independent tasks and track them from one view.
• Knowledge and playbooks — store repo-specific instructions, conventions and recurring procedures so future sessions start with the right context.
• Integrations — connect Slack, Linear, Jira and GitHub so tasks can be created and updated where the team already works.
• Enterprise security — SOC 2, ISO 27001 and FedRAMP posture, per-session VM isolation, RBAC/SSO/SCIM/OIDC, private networking, audit logs, zero data retention with model providers, and IP ownership of inputs and outputs.

What it is good at

Devin fits well when work is well-scoped but tedious: dependency upgrades, test coverage, small features, bug fixes with a reproduction, documentation updates, and migrations across many files. It is less about pair-programming in the editor and more about handing off a queue of tasks and reviewing PRs later.

Getting started

1. Sign up at app.devin.ai and connect a GitHub repository.
2. Create a session and describe the task, linking the issue or PR if you have one.
3. Add repo knowledge (build commands, style rules, test commands) so Devin does not rediscover them each time.
4. Watch the session, answer questions if it gets blocked, and review the resulting PR.
5. For recurring work, save a playbook and reuse it across sessions.

Pricing

Devin is a paid product with subscription/usage plans and enterprise agreements; Cognition offers a demo and trial path. See devin.ai for current plan details.

Links

• Website: devin.ai
• Sign up: app.devin.ai
• Demo: devin.ai/demo

[!warning]
Most of our current development effort is on mini-swe-agent,
which has superseded SWE-agent. It matches the performance performance of SWE-agent, while being
much simpler.
See the FAQ for more details about the differences.
Our general recommendation is to use mini-SWE-agent instead of SWE-agent going forward.

SWE-agent enables your language model of choice (e.g. GPT-4o or Claude Sonnet 4) to autonomously use tools to
fix issues in real GitHub repositories,
find cybersecurity vulnerabilities, or
perform any custom task.

• ✅ State of the art on SWE-bench among open-source projects
• ✅ Free-flowing & generalizable: Leaves maximal agency to the LM
• ✅ Configurable & fully documented: Governed by a single yaml file
• ✅ Made for research: Simple & hackable by design

SWE-agent is built and maintained by researchers from Princeton University and Stanford University.

📣 News

• July 24: Mini-SWE-Agent achieves 65% on SWE-bench verified in 100 lines of python!
• May 2: SWE-agent-LM-32b achieves open-weights SOTA on SWE-bench
• Feb 28: SWE-agent 1.0 + Claude 3.7 is SoTA on SWE-Bench full
• Feb 25: SWE-agent 1.0 + Claude 3.7 is SoTA on SWE-bench verified
• Feb 13: Releasing SWE-agent 1.0: SoTA on SWE-bench light & tons of new features
• Dec 7: An interview with the SWE-agent & SWE-bench team

🚀 Get started!

👉 Try SWE-agent in your browser: (more information)

Read our documentation to learn more:

• Installation
• Hello world from the command line
• Benchmarking on SWE-bench
• Frequently Asked Questions

SWE-agent for offensive cybersecurity (EnIGMA)

SWE-agent: EnIGMA is a mode for solving offensive cybersecurity (capture the flag) challenges.
EnIGMA achieves state-of-the-art results on multiple cybersecurity benchmarks (see leaderboard).
Please use SWE-agent 0.7 while we update EnIGMA for 1.0.

In addition, you might be interested in our other projects:

Contributions

If you'd like to contribute to the codebase, we welcome issues and pull requests! For larger code changes, we always encourage discussion in issues first.

Citation & contact

SWE-agent is an academic project started at Princeton University by John Yang*, Carlos E. Jimenez*, Alexander Wettig, Kilian Lieret, Shunyu Yao, Karthik Narasimhan, and Ofir Press.
Contact person: John Yang, Carlos E. Jimenez, and Kilian Lieret (Email: johnby@stanford.edu, carlosej@cs.princeton.edu, kl5675@princeton.edu).

If you found this work helpful, please consider citing it using the following:

@inproceedings{yang2024sweagent,
  title={{SWE}-agent: Agent-Computer Interfaces Enable Automated Software Engineering},
  author={John Yang and Carlos E Jimenez and Alexander Wettig and Kilian Lieret and Shunyu Yao and Karthik R Narasimhan and Ofir Press},
  booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
  year={2024},
  url={https://arxiv.org/abs/2405.15793}
}

If you used the summarizer, interactive commands or the offensive cybersecurity capabilities in SWE-agent, please also consider citing:

@misc{abramovich2024enigmaenhancedinteractivegenerative,
      title={EnIGMA: Enhanced Interactive Generative Model Agent for CTF Challenges},
      author={Talor Abramovich and Meet Udeshi and Minghao Shao and Kilian Lieret and Haoran Xi and Kimberly Milner and Sofija Jancheska and John Yang and Carlos E. Jimenez and Farshad Khorrami and Prashanth Krishnamurthy and Brendan Dolan-Gavitt and Muhammad Shafique and Karthik Narasimhan and Ramesh Karri and Ofir Press},
      year={2024},
      eprint={2409.16165},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2409.16165},
}

🪪 License

MIT. Check LICENSE.