Liquid AI Releases LFM2.5-VL-3B: A 3B Vision-Language Model That Reads Screens, Grounds Objects, and Calls Tools On-Device

Liquid AI has introduced LFM2.5-VL-3B, a 3.1 billion-parameter vision-language model designed to run locally on devices using approximately 3 GB of memory. The model demonstrates improved performance in screen navigation, object grounding, and tool execution, scoring significantly higher than its predecessor on benchmarks like ScreenSpot-v2 and ToolSandbox. This release enables more complex multimodal capabilities, such as visual analysis and software interaction, to function without requiring cloud-based processing.