llama.cpp Hits 100K GitHub Stars — Gerganov Predicts 90% of AI Agents Running Locally Within 6 Months

Georgi Gerganov / GitHubvia @ggerganov·medium signal

Georgi Gerganov marked llama.cpp's 100K-star milestone with a reflection and a pointed joke: 'Now that 90% of the code worldwide is being written by AI agents, I predict that within 3-6 months, 90% of all AI agents will be running locally with llama.cpp.' He calls 2026 one of the most important years for the local AI movement. Separately, a new Apple Neural Engine (ANE) backend PR targets NPU inference on all Apple Silicon, not just M5's Neural Accelerator cores.