Community Radar.

“With HuggingFace, Nvidia is also acquiring llama.cpp and the team behind it”

With this move Nvidia is not only acquiring the HuggingFace platform, but they might also effectively acquire the copyright to the `llama.cpp` project, together with the entire team behind it. In February 2026 the llama.cpp team was employed by HF in order to continue working on llama.cpp and the ggml library. This includes: - Georgi Gerganov - Xuan-Son Nguyen - Aleksander Grygier - Victor Mustar - Lysandre - Julien Chaumond Now with the acquisition, llama.cpp's future looks a lot less certain given Nvidia's poor track record with open-source. This is still rather speculative at this stage, but it's definitely possible for the llama.cpp project to change in the future: either by switching to a different license, or by having staff redirected to other projects within the larger company. Even when a project is open-source the copyright owner has complete control over it, and they can change licensing as they wish. This has happened before with projects like Redis, Minio, and others. Source: https://huggingface.co/blog/ggml-joins-hf Edit: The original announcement from Feb 2026 from Gerganov gives a few more details: https://github.com/ggml-org/llama.cpp/discussions/19759