community-verified · measured pool data

Can the RTX 3090 24GB ×2 run gemma-4-12b-it-GGUF?

Yes — community-measured at 48.7 tok/s (3 runs at 8-bit).

Verified single-stream runs on the RTX 3090 24GB ×2 from the localmaxxing public pool. Every number on this page is measured — nothing here is estimated.

Quantization	Median tok/s	TTFT (ms)	Peak VRAM	Max context tested	Engines
8-bit	48.7	349.4	—	131,072	llama.cpp

Go deeper

I have a goal — find models that fit my use case → I have hardware — browse everything my rig can run → All machine × model answer pages →