AI Grant

AI Grant is a grant-giving initiative focused on the open source side of artificial intelligence. Rather than investing in companies for equity, it hands out individual awards between $5,000 and $50,000 to people building freely available tools, datasets, libraries and research artifacts. Funding can arrive as cash, as compute credits, or as a mix of the two, depending on what the recipient needs most.

Over the years the program has supported a wide range of work: a Python binding for the llama.cpp inference engine, the GGUF model file format that became a de facto standard for local model distribution, fast cross-platform deep learning libraries for the browser, datasets for sound event research, tools for simulating quantum systems and gravitational lensing, medical imaging work such as stroke scan interpretation and child abuse detection in X-rays, and many smaller research and infrastructure efforts. This breadth reflects the program's interest in useful, reusable building blocks rather than a single vertical.

AI Grant was founded in 2017 by Nat Friedman and Daniel Gross. Note that aigrant.org handles the grant program for open source projects, while the founders' separate accelerator for AI startups lives at aigrant.com. AI Grant operates as an independent organization and is not affiliated with OpenAI, Google or any other model provider named in project descriptions.