A new report from European venture capital firm Runa Capital has shed light on the top 20 trending open source startups, with a significant majority of them closely aligned with artificial intelligence (AI). The report, which is the handiwork of Runa Capital's Runa Open Source Startup (ROSS) Index, showcases the fastest-growing projects in terms of GitHub "stars" – a metric akin to a "like" on social media.

The ROSS Index, which has been operational since 2020, serves quarterly updates on the fastest-growing projects. However, beginning in 2023, Runa started producing annual reports, highlighting the most popular commercial open source startups in a given year. Last year's report demonstrated that AI and data infrastructure were driving demand for open source tooling, with LangChain hitting pole position in the ROSS Index for its open source framework for building LLM-centric apps.

This year's report tells a similar story, with AI central to 11 of the top 20 companies. The top spot on the 2024 ROSS Index is held by Ollama, a Y Combinator alum that's built an open source tool for running LLMs such as Meta's Llama and DeepSeek locally. Ollama's GitHub star count increased by some 76,000 through 2024, growing 261% to more than 105,000. Next on the list is Zed Industries, a cross-platform collaborative code editor "designed for high-performance collaboration with humans and AI."

The report also highlights LangGenius, the company behind an open source LLM app development platform called Dify, which attained more than 43,000 new GitHub stars last year, growing 326% from around 13,000 to nearly 57,000. ComfyUI, an open source node-based program for generating images, videos, and audio using generative AI models, rounds out the top five, with its GitHub star count growing 195% to 61,900 stars last year.

While the ROSS Index for last year illustrates the explosive growth in AI and LLMs, it also shows how developer tooling is still hot in the world of open source, with the likes of Zed and Astral's UV (No. 9) both featuring in the top 10. Elsewhere, the presence of PDF manipulation tool Stirling PDF (No. 7), finance management software Maybe Finance (No. 8), and remote desktop software RustDesk (No. 17) suggests privacy-focused self-hostable tooling is still in high demand.

The report also reveals that San Francisco is home to six of the top 20 ROSS startups, while Canada has three, and Europe (U.K., Switzerland, Hungary, and Czech Republic), Singapore, and China constituting the rest. This highlights the global nature of open source software development, with contributors from all corners of the globe getting involved.

It's worth noting that the ROSS Index is heavily curated and doesn't include any old open source project. Qualifying projects must be closely linked to a commercial company (i.e., a vendor-led project), meaning no side projects. Additionally, these companies must be younger than 10 years old; raised less than $100 million in funding; and be entirely independent — so not a subsidiary or publicly listed.

The methodology behind the ROSS Index is also noteworthy. GitHub "stars" can be an imperfect metric, as it merely shows that someone has "liked" the project, as opposed to actively using or monitoring it. Older projects will naturally have procured more "stars," too, which is why Runa focuses on the relative growth of repositories over a 90-day period for its quarterly reports, and on the absolute number of new stars gained during the year for its annual report.

There may also be some issues around what is classed as "open source." While many of the projects on the list have indeed been released under a recognized copyleft or permissive open source license, this isn't a strict stipulation of the ROSS Index. Runa says it adheres to the "commercial perception" of open source, rather than the official open source definition.

Despite these limitations, the ROSS Index remains a useful indicator not only of what kinds of open source technology are trending but also what companies are trying to build businesses atop them. As the AI-centric startups continue to dominate the list, it will be interesting to see how this trend evolves in the coming years.