Gemini 2.0 Flash Experimental is a beta version model available on Google’s “Ai Studio” (https://aistudio.google.com/): multimodal, it accepts several pdf files, images, videos and audios as input.

In generation, it manages text and image simultaneously and is able to modify the generated images.

The instructions can be edited, the texts can be exported in markdown and the images in jpg.

Free and unlimited.

Tests performed

• Analysis of pdf (source: UNESCO AI Education Repository):

• Audio analysis (source: Extraclasse, parlons paratiques n°38)

• Video analysis (source: AI in 90s):

• Image generation and editing

• Generating an illustrated short story