AI RESEARCH
jina-vlm: Small Multilingual Vision Language Model
arXiv CS.CL
•
ArXi:2512.04032v3 Announce Type: replace We present jina-vlm, a token-efficient 2.4B parameter vision-language model that achieves state-of-the-art multilingual VQA performance among open 2B-scale VLMs. The model couples a SigLIP2 vision encoder with a Qwen3 language decoder and makes use of image tiling and attention-pooling for token-efficient processing of arbitrary-resolution images. To understand the contribution of different