lllm3090#
Local LLM serving for a single RTX 3090, with a browser control panel.
A llama.cpp engine, a web UI on loopback that starts and stops it and downloads models, and a curated model list where every entry has been checked to fit 24 GB with a usable context left over.
The engine exposes both the OpenAI API and Anthropic’s /v1/messages, so
Claude Code and OpenAI-compatible clients work against it without a translation
proxy.
Tutorials
Start here. Install the stack and serve your first model.
How-to guides
Recipes for specific jobs — Claude Code, remote access, your own GGUF files.
Reference
The CLI, the HTTP API, the model catalogue and its fields.
Explanations
Why the project is scoped to one GPU, and what actually decides whether a model fits.