Skip to content

Can Mac M1 8G run LFM3-VL 2B?

Comfortable: GGUF Q8_0 needs 2.0 GB at 4K, batch 1; headroom 4.0 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_M1.2ComfortableVerified file
GGUF Q5_K_M1.4ComfortableVerified file
GGUF Q6_K1.6ComfortableVerified file
GGUF Q8_02.0ComfortableVerified file

Context capacity · GGUF Q8_0

ContextTotal GBVerdict
4K2.0Comfortable
16K2.1Comfortable
32K2.4Comfortable

Generation speed

GGUF Q8_0: 39 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs