#local-llm
Local LLMs on GitHub, November 2–8, 2026
Upstar ranked 109 repos in Local LLMs the week of November 2; together they gained 21,127 stars, 39% more than the week before. synchopate/llama-modchip rose from #18 to #2 with 411 stars gained. The top 10 had 8 repos that were not in it the week before, among them FareedKhan-dev/kimi-k3-in-c at #3, in its first week in the topic. The week brought 34 repos new to the topic's ranking, among them orailnoor/cross-platform-llm-client at #5.
The ranking as the week’s last hourly run left it, on November 9 at 00:00 UTC. It does not change.
- repos ranked
- 109
- stars gained +39% vs the week before
- 21,127
- new in the top 10
- 8 of 10
- ranked in #local-llm for the first time
- 34
The top 10
Ordered by stars gained over the seven days relative to each repo’s size, which is why a repo can rank above one that gained more.
- 1shanselman/gputaskman A GPU task manager for Windows, Linux, and WSL: per-process VRAM, GPU activity, and LLM model/KV-cache insights for NVIDIA, AMD, and Intel GPUs (plus Huawei Ascend on Linux).stars that week +1k
- 2synchopate/llama-modchip Run 284B DeepSeek-V4-Flash (<59GiB RAM: 41.1 cold pp, 9.1 N=1 tg, up to 25.3 spec tok/s) & 30B MoEs (22.4 N=1 / 111.6 spec tok/s) on 64GB CPU (0 GPUs).stars that week +411
- 3FareedKhan-dev/kimi-k3-in-c 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.stars that week +2.1k
- 4vladimirrott/mojo-dllm Local inference for diffusion language models (LLaDA, Dream), written in Mojostars that week +285
- 5orailnoor/cross-platform-llm-client A unified cross-platform AI client supporting seamless transitions between standard cloud APIs and on-device, offline execution of custom and uncensored language models.stars that week +622
- 7RustingStudio/RustingStrata Pure-Rust inference engine for Qwen GGUF models (Qwen3 to Qwen3.8, dense and MoE up to 770B+). CUDA and Vulkan (NVIDIA, AMD), hybrid CPU+GPU experts, MTP speculative decoding, OpenAI and Anthropic compatible server. Up to ~1.7x llama.cpp decode on a 12 GB RTX 3060.stars that week +134
- 8freriadoc/Avalon Fast fork of llama.cpp designed for Qwen3.8-Flash-Nextstars that week +185
- 9Magnvs7/Promethevs PROMETHEVS: run the full DeepSeek-V4.1-Flash-1017 (510 GB) on a gaming PC with 64 GB RAM and a 12 GB GPU. PROMETHEVS engine (compress-then-correct), OpenAI-compatible local server, report and logs.stars that week +125
5 repos of this ranking were removed since and are not shown: deleted from GitHub or taken out by Upstar’s review. Their ranks stay empty.
Ranks 11 to 30
- 11FlashML-org/FreeTokennot ranked the week before +1.2k stars that week
- 12Generated-Computers/Localsup from #25 the week before +200 stars that week
- 13open-jarvis/OpenJarvisup from #16 the week before +1.1k stars that week
- 14lemonadesaltbagel/airlockup from #24 the week before +90 stars that week
- 15lawrence3699/jev-stylenot ranked the week before +84 stars that week
- 16wici-ai/inferssddown from #5 the week before +128 stars that week
- 17anonnerdfrenzy/survival-llmdown from #1 the week before +290 stars that week
- 19simplexai-labs/PolyStratadown from #11 the week before +177 stars that week
- 20mw00/project-mayadown from #10 the week before +303 stars that week
- 21JakeATX/llamAmpereup from #57 the week before +88 stars that week
- 22Dharshaneshwaran/hyperpodnot ranked the week before +55 stars that week
- 23techjarves/Portable-Local-Studionot ranked the week before +209 stars that week
- 24sgaofen/humanizer-local-modelnot ranked the week before +127 stars that week
- 25nace-ai/kapitsanot ranked the week before +116 stars that week
- 26affifuddin-hazam/ghost-infernot ranked the week before +46 stars that week
- 28prabanta-dev/janasdown from #26 the week before +105 stars that week
- 29MakazhanAlpamys/Soupup from #41 the week before +426 stars that week
- 30drawthingsai/local-code-communitydown from #14 the week before +42 stars that week
Show the other 78 repos
- 31johnpsasser/lifeboatnot ranked the week before +42 stars that week
- 32FrigadeHQ/hobdown from #2 the week before +87 stars that week
- 33DevDock-AI/Speedynot ranked the week before +35 stars that week
- 34Lowaltitudeflight/local-llm-tuning-guidenot ranked the week before +33 stars that week
- 35ahtx/kindidown from #22 the week before +34 stars that week
- 36incoai/splashnot ranked the week before +138 stars that week
- 37syv-ai/HyperQwendown from #21 the week before +144 stars that week
- 38davide-troiani/moeshiftnot ranked the week before +28 stars that week
- 39Soulmate-Halo/heterogeneous-gpu-pd-labdown from #33 the week before +90 stars that week
- 40gufo-org/gufonot ranked the week before +88 stars that week
- 41sola1039/HAST-SECdown from #31 the week before +60 stars that week
- 42drumih/turbo-fieldfareup from #84 the week before +218 stars that week
- 43Roy3838/Observernot ranked the week before +102 stars that week
- 44Jumbomuffin777/Frontier-UInot ranked the week before +25 stars that week
- 45lyogavin/airllmnot ranked the week before +492 stars that week
- 46benmaster82/Pithosnot ranked the week before +24 stars that week
- 47Indras-Mirror/Xolotldown from #27 the week before +48 stars that week
- 48peonist-ai/halogen-flash-serverdown from #45 the week before +78 stars that week
- 49NVIDIA/hyperflare-inferencedown from #39 the week before +58 stars that week
- 50ddalcu/mlx-servedown from #29 the week before +101 stars that week
- 51ggml-org/llama.cppnot ranked the week before +742 stars that week
- 52ggml-org/Llama-Windowsdown from #34 the week before +37 stars that week
- 53ollama/ollamadown from #51 the week before +845 stars that week
- 54Wassimyounes01/qwen38-uncensorednot ranked the week before +38 stars that week
- 55wenxuanzhang1209-cyber/meetdonot ranked the week before +30 stars that week
- 56a-ghorbani/pocketpal-aiup from #59 the week before +156 stars that week
- 57janhq/janup from #83 the week before +343 stars that week
- 58Neywa-Labs/mlx-vlmup from #88 the week before +117 stars that week
- 59StartLuxLabs/StartLux-Decisiondown from #54 the week before +61 stars that week
- 60nicedreamzapp/claude-code-localup from #106 the week before +87 stars that week
- 61lemonade-sdk/lemonadenot ranked the week before +113 stars that week
- 62jaredpalmer/kevdown from #42 the week before +141 stars that week
- 63cactus-compute/cactusup from #65 the week before +163 stars that week
- 65jundot/omlxdown from #56 the week before +197 stars that week
- 66AlexsJones/llmfitup from #75 the week before +246 stars that week
- 67trymirai/uzunot ranked the week before +69 stars that week
- 68T-Crypt/berthnot ranked the week before +20 stars that week
- 69AtomicBot-ai/Atomic-Chatdown from #55 the week before +51 stars that week
- 70magnitudedev/magnitudedown from #30 the week before +100 stars that week
- 71cobanov/PS5LMdown from #36 the week before +32 stars that week
- 72unslothai/unslothnot ranked the week before +320 stars that week
- 73jegly/Boxup from #87 the week before +34 stars that week
- 74Mintplex-Labs/anything-llmdown from #66 the week before +273 stars that week
- 75midudev/runonweb.aidown from #28 the week before +28 stars that week
- 76NightMean/OlliteRTup from #79 the week before +23 stars that week
- 77XHToken/Spark-X2.6down from #69 the week before +27 stars that week
- 78alibaba/MNNnot ranked the week before +110 stars that week
- 79ggml-org/Llama-macOSdown from #72 the week before +35 stars that week
- 80microsoft/foundry-localnot ranked the week before +40 stars that week
- 81vllm-project/vllm-metalup from #82 the week before +34 stars that week
- 82mostlygeek/llama-swapdown from #74 the week before +60 stars that week
- 83ml-explore/mlx-lmup from #86 the week before +64 stars that week
- 84Osmantic/ODSup from #94 the week before +63 stars that week
- 85StayLameBro/backburnerdown from #80 the week before +24 stars that week
- 86off-grid-ai/OGAMdown from #85 the week before +40 stars that week
- 87mudler/LocalAInot ranked the week before +150 stars that week
- 88isair/jarvisdown from #81 the week before +31 stars that week
- 89Arthur-Ficial/apfelnot ranked the week before +49 stars that week
- 90exo-explore/exoup from #95 the week before +124 stars that week
- 91google-ai-edge/LiteRT-LMnot ranked the week before +46 stars that week
- 92osaurus-ai/osaurusup from #107 the week before +50 stars that week
- 93mflux-community/mfluxnot ranked the week before +25 stars that week
- 94google-ai-edge/galleryup from #101 the week before +77 stars that week
- 95youssofal/MTPLXdown from #89 the week before +24 stars that week
- 96chatboxai/chatboxsame rank as the week before +94 stars that week
- 97ikawrakow/ik_llama.cppup from #105 the week before +25 stars that week
- 98raullenchai/Rapid-MLXdown from #77 the week before +27 stars that week
- 99AtomicBot-ai/atomic-agentdown from #73 the week before +22 stars that week
- 100OpenBMB/MiniCPM-Vdown from #99 the week before +59 stars that week
- 101PrismML-Eng/Bonsai-demodown from #91 the week before +20 stars that week
- 102mozilla-ai/llamafileup from #104 the week before +49 stars that week
- 103Andyyyy64/whichllmup from #109 the week before +23 stars that week
- 104qualcomm/GenieXup from #108 the week before +25 stars that week
- 105cactus-compute/enginenot ranked the week before +20 stars that week
- 106n4ze3m/page-assistup from #110 the week before +22 stars that week
- 107deepseek-ai/DeepSeek-R1not ranked the week before +63 stars that week
- 108mlc-ai/web-llmnot ranked the week before +23 stars that week
- 109oobabooga/textgenup from #112 the week before +23 stars that week