Multimodal model

VisionPsy Nano 460M

VisionPsy-Nano-460M is a compact ~460M parameter vision-language model built for on-device and edge deployment. Built on the nanoVLM architecture with a SigLIP2 vision encoder and SmolLM2 language backbone, it handles visual question answering, document and chart understanding, and scene-text reading. Despite its small size, it compares favorably to similarly sized vision-language models on 16 of 17 public benchmarks and outperforms models up to twice its size on reasoning and perception tasks.