8 local LLM settings most people never touch that fixed my worst AI problems:

xda-developers.com/moe-models-changed-what-fits-on-my-gpu-means-and-most-hardware-advice-hasnt-caught-up/

Credibility C

Oct 28

MoE models changed what "fits on my GPU" means, and most hardware advice hasn't caught up

github.com/madhusoodhans/model-cache-cleaner

github.com Oct 26

GitHub - madhusoodhans/model-cache-cleaner: Claude skill that finds and safely cleans storage used by local AI tools (LM Studio, Ollama, Hugging Face ...

xda-developers.com/ollama-easiest-way-start-with-local-ai-its-also-first-thing-id-replace/

Oct 24

Ollama is the easiest way to start with local AI, and it's also the first thing I'd replace

2 similar articles

github.com Oct 07

GitHub - Chida82/sf-ds4-1flash: Contribute to Chida82/sf-ds4-1flash development by creating an account on GitHub.

github.com/Chida82/sf-ds4-1flash

xda-developers.com/local-llm-wasting-ram-context-never-use-one-setting-gives-back/

Oct 05

Your local LLM is quietly wasting RAM on context you never use, and one setting gives it back

xda-developers.com/stopped-my-local-llm-agent-from-running-out-of-context/

Sep 27

I stopped my local LLM agent from running out of context, and now it handles what Claude Code does

xda-developers.com/stopped-downloading-bigger-local-models-started-running-two-small-ones-at-once/

Sep 25

I stopped downloading bigger local models and started running two small ones at once

xda-developers.com/my-local-ai-model-was-wasting-vram-by-default-and-lowering-one-setting-doubled-its-speed/

Sep 25

My local AI model was wasting VRAM by default, and changing one setting doubled its speed

xda-developers.com/x-numbers-i-check-before-downloading-a-local-model-and-tokens-per-second-isnt-the-important-one/

Sep 24

I check these 4 numbers before downloading a local AI model, and tokens per second is dead last

xda-developers.com/settings-i-always-change-before-running-local-llm/

Jul 02

6 settings I always change before running a local LLM: You might not need a different model, but better settings

github.com/razvanneculai/litecode

github.com Apr 28

GitHub - razvanneculai/litecode: CLI Tool for 8k context ai models: CLI Tool for 8k context ai models. Contribute to razvanneculai/litecode development by ...

github.com/ParmesanParty/llama.cpp

github.com Apr 15

GitHub - ParmesanParty/llama.cpp: Specialized fork for (relatively) fast single-GPU inference (in CUDA) using large MoE models that don't fit fully into ...

github.com/peva3/SmarterRouter

github.com Feb 20

GitHub - peva3/SmarterRouter: SmarterRouter: An intelligent LLM gateway and VRAM-aware router for Ollama, llama.cpp, and OpenAI. Features semantic caching ...

github.com/elemein/agent-stoat

github.com Feb 20

GitHub - elemein/agent-stoat: Local-first coding agent built with small models in mind and tight VRAM budgets.