Welcome
Work locally with deployed open-models in your cloud.
llmrun runs open-source LLMs on AWS GPU instances and exposes them as a local OpenAI-compatible endpoint. No local GPU, no public IP, no per-token billing.
Getting Started
Install, verify, and spin up your first model in minutes.
Configuration
llmrun.yaml catalog, defaults, and the resolution order.
Model Catalog
VRAM sizing, quantization, gated models, and multi-GPU.
Studio
Local dashboard: live metrics, logs, chat, and cost history.
Coding Assistants
Wire up Continue, Cursor, Cline, Kilo Code, Aider, and more.
Troubleshooting
Common issues encountered in the field, with actionable fixes.