Georgi GerganovA

Gerganov created llama.cpp, enabling LLMs to run on consumer CPUs without a GPU, fundamentally democratizing local AI. His ggml library and GGUF format became the standard for offline inference, powering Ollama, LM Studio, and dozens of tools. Made an entirely new category of private, on-device AI possible. Tone: code-speaks-for-itself, minimal self-promotion, GitHub-centric, lets community build on his work, quietly transformative.