Liquid AI released an experimental DSpark draft model for its vision-language model LFM2.5-VL-3B. It uses speculative decoding to speed up inference without changing output quality, achieving up to 3.13x decode speedup on device and 2.66x on H100, with end-to-end gains up to 2.62x and 2.27x. The drafter adds 280M parameters, an 8.9% increase, and supports llama.cpp, MLX-VLM, and SGLang from day one.