Current AI Unveils a Map of the Gaps in Open AI

The nonprofit Current AI, which describes itself as “a global partnership to build a public-interest alternative in the field of AI,” has unveiled the Gap Map — a map that indexes the current state of open artificial intelligence.

Current AI was founded at the AI Action Summit in Paris in February 2025 and has already attracted around 400 million dollars in investment. Version 0.1 of the Gap Map was launched a few days ago.

What the map covers

The Gap Map describes 421 products in detail: 266 software tools and libraries, 85 models, 50 datasets, and 20 hardware projects. These products were created by 228 organizations.

They are all distributed across 14 categories within three levels of the stack: model components, products and user experience, and infrastructure.

Beyond the products described in detail, the project tracks another 24,400 artifacts — the so-called “long tail” of the open AI ecosystem. These entries have not yet been assessed and will be classified as the research progresses.

Open data

The map itself is interesting to explore, but the data underlying it is no less valuable. It has been published under the MIT license in the currentai-org/os-ai-map repository on GitHub and includes 1,184 YAML files, as well as the notebooks, schemas, and scripts used to gather the information.

Because the files are hosted on GitHub, they can be explored using Datasette Lite. In particular, a CSV file is available listing the 16,185 GitHub repositories the project keeps an eye on — it, too, can be loaded directly into Datasette Lite for analysis.

Source: Simon Willison