#1
Ranked by stars: 8,534
deepseek-ai/DeepGEMM
DeepGEMM: clean and efficient BLAS kernel library on GPU
8534 stars · 1361 forks · Cuda
Cuda repositories
Ordered by stars within our imported repository sample.
Snapshot: 2026-10-31T12:18:20.650Z. Rankings use this list’s declared public metric; missing values sort last, ties use record ID, and Trending retains source order.
#2
Ranked by stars: 2,306
mw00/project-maya
Project Maya - run GLM-5.3-Flash (321B MoE), DeepSeek-V4.1-Flash-1017 (748B MoE) and MiniMax-M3.1-Flash (428B MoE) privately on your own GPU(s), from one card to 16, NVIDIA or AMD, Linux or Windows. Engine, server, dashboard, OpenAI- and Anthropic-compatible API, quant tools.
2306 stars · 310 forks · Cuda
#3
Ranked by stars: 357
deepseek-ai/DeepSelect
DeepSelect: TopK kernels for DeepSeek Sparse Attention (DSA) and Samplers
357 stars · 23 forks · Cuda