LFM2.5-VL-3B: A Better and Faster Vision-Language Model for the Edge

Initial benchmarks on my M1 Pro, about 10x faster at describing screenshots than qwen3.5-4b. If the quality is comparably, as the model sheet implies, that’s incredible. But it seems this isn’t a reasoning model?

I worked on this model at Liquid and will be checking in to answer any questions/discussions, both about this vision model and LiquidAI more broadly.