Skip to main content
← Back to Vision Models LFM2.5-VL-450M is Liquid AI’s smallest vision-language model. It is a refreshed version of LFM2-VL-450M, built on the LFM2.5-350M language backbone with an 86M SigLIP2 NaFlex vision encoder. Compared to LFM2-VL-450M, it improves instruction following and multilingual vision understanding, and adds bounding box prediction and function calling (text-only input).

Specifications

Ultra-Light

Minimal memory footprint

Low Latency

Fastest vision model inference

Edge-Ready

Runs on mobile and embedded devices

Quick Start

Last modified on November 17, 2026