Gemini and the multimodal turn

Google's answer to GPT: a family of models built from the beginning to handle text, images, audio, and video.

Gemini(Model), built by Google(Company), is a family of multimodal models.

Design

Unlike models that add vision as an afterthought, Gemini was pre-trained on text, images, audio, and video jointly.

Entities mentioned

Continue exploring the forest

Next trail→Google

Every article leads somewhere. Follow this entity, or search the whole forest.