MODEL-RUNNER / AIRLLM
AirLLM
Run 70B models on 4GB RAM via layer-wise inference.
model-runnerfreelocal
locallow-ram
Kindmodel-runner
Accesslocal
Pricingfree
Free tier—
Caps2
GitHub ★34,534as of 2026-09-19
MACHINE-READABLE
GET /api/v1/resources/airllmOpen JSON ↗