What Can I Run AI?

Comparisons for local LLMs

Stuck between two cards, or two models? Hardware pages show which open models each machine runs at Q4. Model pages show memory, context cache, cheapest hardware and estimated speed side by side.

Model vs model

GPU vs GPU

Or see buying guides by model family

New models, and what they need to run

One short email a week: fresh open models with their VRAM numbers, plus hardware worth watching.

Or subscribe here