Skip to main content
Google’s Gemini models in DeepMask are all from the Gemini 3 Flash family. They are built for speed and share a 1M-token context window, up to 65K tokens of output, image and document input, and tool use. The Flash models are the stronger generalists; the Flash Lite models trade some depth for speed and lower cost.

About

Gemini 3.8 Flash is the newest Gemini model in DeepMask. Its 1M-token context window makes it a good fit for reading large document sets, long transcripts, or many images in one conversation.

Specifications

Google’s open-weight Gemma models are covered on the Gemma page.
Try Gemini in DeepMask →