PaliGemma – Google's Cutting-Edge Open Vision Language Model

In three linesGoogle releases PaliGemma, an open-source vision-language model built on Gemma 2B. The model combines an image encoder and text decoder for multilingual visual understanding tasks. Weights available on Hugging Face.

Summary generated by Claude — human-verified