DeerFlow vs Open SWE
Which should you use? A side-by-side look at what each one does, who it suits, what it costs, how hard it is to set up, and where each falls short.
Open source · MIT
An open-source agent framework that coordinates sub-agents, memory, and sandboxes for long, multi-step tasks.
Open source · MIT
Open-source coding agent that takes engineering tasks and delivers pull requests, meant to be self-hosted.
The short answer
Choose DeerFlow if you need…
- Developers who want to experiment with multi-agent workflows on their own infrastructure
- Researchers running long, multi-step investigations that need memory and sub-agents
- Teams already using LangGraph or LangChain who want a reference agent harness
- Users who are comfortable configuring their own LLM API keys and running services locally or in Docker
Choose Open SWE if you need…
- Engineering teams that want a self-hosted coding agent they can inspect and adapt
- Developers who already run LangChain or LangGraph and want a reference implementation
- Teams that want an agent to handle routine PR work such as reviews and CI follow-ups
- Organizations that need control over sandbox providers, models, and GitHub access
Side-by-side
| DeerFlow | Open SWE | |
|---|---|---|
| Type | AI model / agent | AI model / agent |
| Price | Free and open source | Free and open source |
| Licence | MIT | MIT |
| GitHub stars | 84.9k | 11.2k |
| Setup difficulty | Advanced | Advanced |
| How you get it | Run with Docker | Clone and build from source |
| Main language | Python | Python |
| Last code update | 4 days ago | 4 days ago |
What each one is
DeerFlow
DeerFlow is an open-source 'super agent harness' that runs an AI lead agent which can split work across sub-agents, keep long-term memory, and execute code or tasks inside sandboxes. It is built to handle tasks that may take minutes to hours, such as research, coding, and content creation. The README describes it as extensible through skills and tools.
Open SWE
Open SWE is an open-source coding agent from LangChain that investigates a codebase, makes changes in an isolated sandbox, validates them, and opens a pull request. It can also review pull requests, monitor CI, and respond to feedback. It is built on Deep Agents and LangGraph and is designed to be deployed in your own infrastructure.
Best for
DeerFlow
- Developers who want to experiment with multi-agent workflows on their own infrastructure
- Researchers running long, multi-step investigations that need memory and sub-agents
- Teams already using LangGraph or LangChain who want a reference agent harness
- Users who are comfortable configuring their own LLM API keys and running services locally or in Docker
Open SWE
- Engineering teams that want a self-hosted coding agent they can inspect and adapt
- Developers who already run LangChain or LangGraph and want a reference implementation
- Teams that want an agent to handle routine PR work such as reviews and CI follow-ups
- Organizations that need control over sandbox providers, models, and GitHub access
Strengths
DeerFlow
- Open-source under the MIT license, so the code can be inspected and modified
- Combines sub-agents, memory, sandboxes, and skills in one harness
- Documented deployment options including Docker and local development
Open SWE
- Covers the full loop from investigation to pull request, review, and CI follow-up
- Open source under MIT, and deployable in your own infrastructure
- Lets you choose models, sandbox providers, integrations, and approval policies
Not ideal for
DeerFlow
- Non-technical users looking for a ready-to-use consumer app, since setup involves configuration, Docker or local development, and API keys
- Anyone who needs a fully managed, hosted service with no operational work
- Security-sensitive deployments without reviewing the README's security notice, because the gateway admin role is equivalent to code execution
Open SWE
- Anyone wanting a ready-to-use hosted product: the README says deploying as-is requires Managed Deep Agents, otherwise you must fork and adapt the code to your own runtime
- Teams that expect stable releases: the README says it is under active development with expected breaking changes
- People hoping to contribute issues or patches: the project is not accepting issues or external contributions at this time
- Anyone who needs strong isolation for local execution: the README says local execution does not provide cloud sandbox isolation
Trade-offs
DeerFlow
- Setup requires multiple prerequisites and configuration, which the README describes as non-trivial
- The README includes a security notice warning that improper deployment can introduce risks
- The README is long and truncated here, so some details of setup and behavior could not be verified
Open SWE
- Under active development with breaking changes and no stability guarantee
- Deployment as-is depends on Managed Deep Agents; otherwise you must adapt it yourself
- Sandboxes have powerful tools and may have network access, so credentials must be tightly scoped
Getting started
DeerFlow
Clone the repository
git clone https://github.com/bytedance/deer-flow.gitOpen SWE
Clone the repository
git clone https://github.com/langchain-ai/open-sweFrequently asked questions
What is the difference between DeerFlow and Open SWE?
DeerFlow: An open-source agent framework that coordinates sub-agents, memory, and sandboxes for long, multi-step tasks. Open SWE: Open-source coding agent that takes engineering tasks and delivers pull requests, meant to be self-hosted. DeerFlow suits developers who want to experiment with multi-agent workflows on their own infrastructure, while Open SWE suits engineering teams that want a self-hosted coding agent they can inspect and adapt.
Are DeerFlow and Open SWE free?
DeerFlow: Free and open source. Open SWE: Free and open source.
Is DeerFlow or Open SWE more popular?
DeerFlow has more GitHub stars (84.9k) than Open SWE (11.2k). Popularity isn't the same as fit — check which one matches your use case above.
What are other alternatives to DeerFlow and Open SWE?
Other options people compare them with include LangGraph, CrewAI, OpenHands, Edict, SWE-agent.
Keep comparing
Built from our catalogue entries for both products: open-source details come from each project's README (AI-assisted, refreshed weekly) and GitHub data; AI-tool details are editorial. Spotted something wrong? Let us know from either product's page.