# Liquid AI — LFM2.5-VL-DSpark: Accelerating vision-language models on edge and beyond

- Company: Liquid AI (liquid.ai)
- Announced: 2026-09-24
- Category: model-update
- Coverage: 1 outlet
- Announcement: yes
- Group: models
- Source: https://www.liquid.ai/blog/lfm2-5-vl-dspark
- Record: https://forck.live/items/16767-lfm2-5-vl-dspark-accelerating-vision-language-models-on-edge-and-beyond
- Subject: LFM / d1 models
- Models affected: LFM2.5-VL-3B

Liquid AI released an experimental DSpark draft model for its vision-language model LFM2.5-VL-3B, achieving decoding throughput improvements of up to 2.66× on GPUs and 3.13× on edge devices without changing output quality. The drafter has approximately 280M parameters, increasing the deployed model’s parameter count by 8.9%. It is available on Hugging Face with support in llama.cpp, SGLang, and MLX-VLM.

## Evidence

Verbatim from https://www.liquid.ai/blog/lfm2-5-vl-dspark:

> We achieve decoding throughput improvements of up to 2.66× on GPUs and 3.13× on edge devices, with end-to-end throughput gains of up to 2.27× and 2.62×, respectively.

## Around this story

Outlets this record can name and link:

- Marktechpost — Liquid AI Releases LFM2.5-VL-3B-DSpark: Speculative Decoding for Vision-Language Models With Up to 3.13x Faster Decoding: https://marktechpost.com/2026/09/25/liquid-ai-releases-lfm2-5-vl-3b-dspark-speculative-decoding-for-vision-language-models-with-up-to-3-13x-faster-decoding

---

Record: https://forck.live/items/16767-lfm2-5-vl-dspark-accelerating-vision-language-models-on-edge-and-beyond
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md