Moondream
Open vision models made for production environments with built-in grounding skills.
Open-source family of vision-language models built for efficient visual reasoning. The official Moondream model page lists multiple variants—Moondream 3 Preview, Moondream 2, and Moondream 2 0.5B—and documents built-in grounded vision skills such as object detection, pointing, captioning, visual question answering, segmentation, and OCR.
Model Intelligence
Context window
32,000 tokens
Benchmarkable
No
Model level
family
Recent stories
0 linked stories
No linked stories yet.