Daniel Bevenius
llama.cpp code owner for model-conversion tooling and member of its CI team; Hugging Face
- Hugging Face is the company listed on Bevenius's GitHub profile, Hugging Face1
- Code owner for model-conversion and the XCFramework build; member of the CI and llama-common teams, llama.cpp project3
Daniel Bevenius contributes to llama.cpp, the C/C++ runtime for running language models on local hardware; Bevenius's GitHub profile lists Hugging Face as company and links to ggml.ai, the company behind the ggml library, which Hugging Face acquired in 2026. llama.cpp's CODEOWNERS file names Bevenius as owner of the model-conversion example and the XCFramework build script, co-owner of the debug example, and a member of the CI and shared-code (llama-common) teams. Bevenius's recent commits cover build-time improvements, API and ABI checks in the release workflow, parallel server tests, and model-conversion tooling.1234
Work
llama.cppOpen-source code
CODEOWNERS names Bevenius for examples/model-conversion, build-xcframework.sh and (with Piotr Wilkin) examples/debug, and lists Bevenius in the ggml-org/ci and ggml-org/llama-common teams. Commits authored in August to October 2026 cover CMake build speedups, an API/ABI check in the release workflow, pytest-xdist for server tests, and model-conversion recipes. The repository is MIT-licensed.345
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.