State of Open Source on Hugging Face: Spring 2026

Hugging Face’s 2026 analysis shows the open‑source AI ecosystem has expanded rapidly, with users, models and datasets nearly doubling in 2025. The platform now hosts 13 million users, more than 2 million public models and over 500 000 public datasets, reflecting a move from passive consumption to active creation of derivative artifacts such as fine‑tuned models, adapters and benchmarks. Despite this growth, the ecosystem remains highly concentrated: about half of all models have fewer than 200 downloads, and the top 200 models—only 0.01% of the catalog—account for 49.6% of all downloads. Geographic dynamics have shifted, with China surpassing the United States in monthly and overall downloads; Chinese models now represent 41% of total downloads. Industry participation fell from roughly 70% before 2022 to about 37% in 2025, while independent developers rose from 17% to 39%, at times exceeding half of total usage. Large firms such as NVIDIA, Meta, Google and Baidu have increased their repository output, with NVIDIA emerging as the strongest contributor. The rise of open‑source models is also tied to sovereignty concerns, as governments seek to fine‑tune systems on local data and reduce reliance on foreign cloud infrastructure. National initiatives, such as South Korea’s Sovereign AI Initiative and Switzerland’s Swiss AI initiative, illustrate how open‑source platforms are becoming strategic tools for domestic AI development. This trend underscores the growing importance of open‑source ecosystems for fostering innovation, reducing costs, and enabling regulatory compliance across the AI industry.