A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
GitHub repository with 4,256 stars and 663 forks.
Language: C
Topics: avx2, c99, cpu-inference, deep-learning, from-scratch, inference-engine, kimi-k3, linear-attention, llm, llm-inference
Trending score 5.42, freshness score 0.79, stars gained +498, forks gained +51.
2026-08-10: 4,256 stars and 663 forks.
DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm
GitHub repository with 21,052 stars and 1,888 forks.
Trending score: 4.88; stars gained: +373; forks gained: +48.
Language: C
Gen1Recomp - A native Lua / LÖVE2D recreation of Gen 1 Poke
GitHub repository with 2,743 stars and 228 forks.
Trending score: 4.42; stars gained: +149; forks gained: +16.
Language: C
Linux kernel source tree
GitHub repository with 242,268 stars and 63,829 forks.
Trending score: 4.41; stars gained: +142; forks gained: +27.
Language: C
High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.
GitHub repository with 38,309 stars and 3,050 forks.
Trending score: 4.18; stars gained: +84; forks gained: +7.
Language: C
Topics: aider, ast, claude-code, code-analysis, code-intelligence, codex
authenticate, sudo, login with your fingerprint wire(less)ly without having to spend $149 at the cost of some security
GitHub repository with 1,243 stars and 59 forks.
Trending score: 3.79; stars gained: +49; forks gained: +5.
Language: C
Run the full 2.78-trillion-parameter Kimi K3 model beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
GitHub repository with 1,983 stars and 145 forks.
Trending score: 3.76; stars gained: +54; forks gained: +2.
Language: C