Google has published the model card for Gemini 3.1 Pro, describing it as the next iteration in the Gemini 3 series and Google’s most advanced model for complex tasks as of February 2026.
The model is based on Gemini 3 Pro and supports multimodal inputs including text, audio, images, video, and code repositories with a 1M token context window. It produces up to 64K tokens of text output and is distributed via channels such as the Gemini App, Google Cloud Vertex AI, and the Gemini API.
Google states that Gemini 3.1 Pro significantly outperforms Gemini 3 Pro across benchmarks requiring enhanced reasoning and multimodal capabilities. The model is particularly suited for agentic performance, advanced coding, and long-context understanding.