Georgi Gerganov
Develops llama.cpp and the ggml tensor library for running models on local hardware
- Maintains llama.cpp with the ggml team, Hugging Face42
Georgi Gerganov develops ggml, a tensor library for machine learning on commodity hardware, and llama.cpp, a C/C++ runtime for running language models locally. Gerganov's site also lists whisper.cpp, which runs OpenAI's Whisper speech recognition model on the CPU. Gerganov founded ggml.ai in 2023 to support ggml development, and Hugging Face acquired the company in 2026. Hugging Face's announcement says the ggml team still spends all of its time maintaining llama.cpp and keeps full leadership of the project's technical direction.134
Work
llama.cpp
Listed among the projects on Gerganov's site and announced by Hugging Face as the ggml team's project. The repository's CODEOWNERS file names Gerganov for core areas including the CPU and Metal backends, quantization code, and llama-server, and Gerganov continues to author commits to it.1456
ggml (external site: github.com)
Founded ggml.ai in 2023 to support development of ggml, the tensor library with integer quantization support that llama.cpp and whisper.cpp are built on. The library is MIT-licensed.3
whisper.cpp (external site: github.com)
Listed on Gerganov's site as inference of OpenAI's Whisper speech recognition model on the CPU, and pinned on Gerganov's GitHub profile as a C/C++ port of the model.12
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.