Open Source AI Agents Ranked by GitHub Stars

Most Starred AI Agent Projects

AutoGen from Microsoft Research has accumulated over 40,000 GitHub stars, making it one of the most starred multi-agent frameworks. Its popularity stems from the Microsoft brand, early-mover advantage in the multi-agent space, and its unique conversational approach to agent collaboration. AutoGen enables multi-agent conversations where agents discuss and iterate on solutions through dialogue. However, its documentation and developer experience have not kept pace with its popularity, and many developers report a steeper learning curve than the star count would suggest.

OpenHands (formerly OpenDevin) has surpassed 70,000 stars with its ambitious vision of fully autonomous software engineering. The project attracted massive attention because it demonstrated agents that could analyze repositories, plan changes, write code, run tests, and iterate. The star growth reflects developer excitement about the potential of autonomous coding rather than widespread production adoption. OpenHands delivers real capability for well-defined tasks but remains a project where promise and reality are converging gradually.

Dify has crossed 80,000 stars as a low-code platform for building AI applications. Its popularity is driven by accessibility, because the visual workflow builder and built-in RAG make AI application development accessible to non-developers. Dify offers genuine production value with its Apache 2.0 license, self-hosting capability, and multi-model support. This is a case where the high star count does correlate with real usability and active adoption.

LangChain and LangGraph together represent the most-starred framework ecosystem with combined stars exceeding 100,000. LangChain was the first major AI agent framework and attracted early adopters across the entire LLM application space. LangGraph, the newer production-focused component, provides the graph-based agent architecture that serious deployments need. The star count reflects the ecosystems foundational role in the AI agent development community.

Hermes Agent reached 60,000 stars within two months of launch, demonstrating the fastest growth trajectory in the AI agent ecosystem. This rapid accumulation reflects strong developer demand for a framework that balances power with simplicity. Whether the star velocity translates to lasting adoption depends on how the project matures in its documentation, stability, and community support over the coming months.

Most Starred Specialized Agents

Browser Use has accumulated significant stars as the leading browser automation framework, with its MIT license, multi-provider model support, and vision-based page understanding driving organic adoption among developers building web automation tools. The star growth correlates with genuine usage because browser automation has concrete, measurable outcomes that drive real adoption.

Aider maintains strong and growing star counts among coding agents, reflecting sustained adoption by developers who use it daily. Unlike projects that get starred and forgotten, Aiders star growth tracks with its release frequency and feature additions. The terminal-based workflow and model flexibility keep developers coming back, which generates organic star growth through word-of-mouth recommendations.

Open WebUI has become one of the most starred self-hosted chat interfaces, driven by demand for privacy-first AI chat alternatives. Its polished interface and seamless Ollama integration make it the default recommendation for anyone looking to run their own ChatGPT alternative, which generates consistent organic star growth from developers and privacy-conscious users alike.

What Stars Actually Measure

GitHub stars primarily measure awareness and initial interest. A star is a single click that takes less than a second and requires no evaluation of the project. Many developers star repositories they find interesting without ever cloning the code, running the software, or evaluating its quality. Viral social media posts, prominent blog articles, and conference talks drive star growth independent of the projects actual capability.

Stars correlate with ecosystem size and marketing effort more than technical quality. Projects backed by large companies (Microsoft, Google) or maintained by developers with large social media followings accumulate stars faster than technically superior projects with less visibility. A project with 5,000 stars and a dedicated community of active users may be more production-ready than a project with 50,000 stars that most people starred and forgot about.

Star velocity (growth rate) can indicate genuine market interest or viral marketing. A project that gains 10,000 stars in a week is probably benefiting from a viral moment rather than organic adoption. A project that gains 10,000 stars over six months while showing steady contributor growth and issue resolution is more likely to have genuine, sustainable adoption. Look at the star growth chart on GitHub to distinguish viral spikes from organic growth.

Stars do not indicate maintenance health. Many highly-starred repositories are abandoned, with no recent commits, unresolved issues, and no release activity. A project with 100,000 stars and no activity for six months is a bigger risk than a project with 5,000 stars and weekly releases. Always check the last commit date, recent release history, and open issue response time before relying on a project.

Metrics That Matter More

Commit frequency in the last 90 days indicates active development. Projects with regular commits are actively maintained, which means bugs get fixed, dependencies get updated, and new features get added. A project that had its last commit three months ago is likely abandoned or deprioritized. Check the commit graph on GitHub to see the development cadence.

Issue response time reveals how the maintainers treat their community. Projects where issues get acknowledged within 24-48 hours and resolved within a reasonable timeframe are well-maintained. Projects where issues sit unanswered for weeks indicate either overwhelmed maintainers or a project that has shifted its priorities away from community support. Try opening a minor issue or asking a question in discussions to test responsiveness before committing to a project.

Contributor count and diversity show project resilience. A project maintained by a single developer is at risk of abandonment if that developer moves on to other work. A project with 50+ contributors spread across multiple organizations is much more likely to survive individual departures. Check the contributor graph to see whether contributions come from many people or concentrate around a single maintainer.

Release cadence and semantic versioning indicate project maturity. Projects that follow semantic versioning (major.minor.patch) and produce regular releases with changelogs take backwards compatibility seriously. Projects that push directly to main with no tagged releases make it difficult to depend on stable versions. A regular release schedule also indicates a structured development process rather than ad-hoc coding.

How to Use These Rankings

Use star counts as a starting filter, not a final decision. If you are exploring a new category of AI agents, sorting by stars gives you a reasonable list of projects to evaluate further. The most-starred projects in each category are at least worth looking at because they have attracted community attention for some reason. But never choose a project based on stars alone.

After identifying candidates through star counts, evaluate each project on the metrics that matter for your specific use case. Check the documentation quality by trying to follow the quickstart guide. Check the maintenance health by reviewing recent commits and issue responses. Check the license compatibility with your intended use case. Check the model support to verify compatibility with your preferred LLM provider. These practical evaluations reveal far more about project quality than star counts ever will.

Be skeptical of recently viral projects with explosive star growth and limited release history. These projects may have real potential, but they have not been tested by real-world deployments. Give them time to mature before depending on them for production use. The projects that deserve the most confidence are those with steady star growth over a year or more, active contributor communities, regular releases, and documented production deployments.