# Liquid AI says LFM2.5-VL-DSpark speeds inference by up to 3.13x

- Published: 2026-09-24T14:08:57.000Z
- Updated: 2026-09-25T18:06:11.945Z
- Section: Development
- Topics: Research
- Page: https://aidailyjournal.com/en/n/2026/09/24/lfm2-5-vl-dspark-speeds-up-vision-language-model-inference/
- Sources: [Hugging Face](https://huggingface.co/blog/LiquidAI/lfm2-5-vl-dspark)

## Summary

Liquid AI released LFM2.5-VL-DSpark, an experimental drafter model for its LFM2.5-VL-3B vision-language model. The company says speculative decoding speeds generation without changing output quality, with gains of up to 3.13x on-device and 2.66x on an H100, while end-to-end latency improves by up to 2.62x and 2.27x, respectively. The drafter adds about 280 million parameters, or 8.9% to the 3-billion-parameter model.

The release includes day-one support for llama.cpp, MLX-VLM and SGLang. Liquid AI evaluated six vision tasks, including image and chart question answering, captioning and complex reasoning. The technique accelerates decoding only; image encoding and prefilling remain unchanged, which can limit the overall gain when those stages account for a large share of response time.

## Why it matters (editorial interpretation)

The technique could reduce generation time for LFM2.5-VL-3B across different execution environments, with day-one support for llama.cpp, MLX-VLM and SGLang. Because it accelerates decoding only, the total benefit may be smaller when image encoding and prefilling account for most of the latency.

_Summary written by AI Daily Journal in its own words, based on the sources above. For details, read the original articles._