Google's family of open vision-language models built on the Gemma architecture. PaliGemma combines image and text understanding, making it great for tasks like image captioning, visual question answering, and document understanding. Optimized for low‑resource environments.