Skip to content

Can RTX 3090 run LFM2.5 24B-A2B?

Comfortable: GGUF Q4_K_M needs 15.3 GB at 4K, batch 1; headroom 8.7 GB.

Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.

Measured runs

No measured run for this pair. Sizing is an estimate, not a runtime test.

Quant memory at 4K

QuantTotal GBVerdictWeight sizing
GGUF Q4_K_M15.3ComfortableVerified file
GGUF Q8_026.7Does not fitVerified file

Context capacity · GGUF Q4_K_M

ContextTotal GBVerdict
4K15.3Comfortable
16K15.5Comfortable
32K15.9Comfortable
128K17.9Comfortable

Generation speed

GGUF Q4_K_M: 664 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs