Google Gemini 3.1 Ultra Has a 2M Token Window

Gemini 3.1 Ultra Is Here

Google has launched Gemini 3.1 Ultra – featuring a massive 2-million token context window that works natively across text, image, audio, and video.

True Multimodal Reasoning

Unlike previous models, Gemini 3.1 Ultra was trained from the ground up to reason across all modalities simultaneously – not just process them separately.

2 Million Tokens Changes Everything

The 2M context window means you can feed entire codebases, books, or hours of video into a single prompt and get coherent analysis back.

Gemma 4 Also Released

Alongside Ultra, Google released Gemma 4 – its most powerful open-source models yet, designed specifically for advanced reasoning and agentic AI workflows.

Competing With Anthropic and OpenAI

This launch comes as Anthropics Model Context Protocol crosses 97M installs and OpenAI prepares for a potential trillion-dollar IPO.

Available Now

Gemini 3.1 Ultra is rolling out across Google products including Search, Workspace, and the Gemini app. Developers can access it via the Gemini API.