AIMarkTechPostabout 2 hours ago
Liquid AI Releases On-Device Vision-Language Model
4 min read
Liquid AI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model for on-device use. It reads screens, grounds objects, parses documents, and calls tools. It averages 69.4 on 28 vision benchmarks, matching InternVL-3.5-4B and trailing Qwen3.5-4B by 0.7 points. It fits in 3 GB and decodes 228 tokens/s on an Apple M5 Max.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
