Google introduced spatial understanding in Gemini 1.5 and they’ve advanced it further with Gemini 2.0. A new tool in Google AI Studio makes it easier to explore spatial understanding with multimodal reasoning. This makes Gemini 2.0 the only major LLM that was trained on spatial 2D and 3D data.
What next?
Watch a demo of spatial understanding: youtube.com/watch?v=-XmoDzDMqj4
Start building with Gemini 2.0 in AI Studio: aistudio.google.com.
If you know of other models with spatial understanding, please comment below with a link.