If you want even better AI models, you might have reason to celebrate. Google announced the release of Gemini 3.1 Pro on Thursday, characterizing the model’s arrival as a “core theoretical advancement.”
As measured by the frequency of machine learning model releases, Gemini 3.1 Pro closely follows recent models debuted from Anthropic and OpenAI. There is hardly enough time to start using new U.S. commercial AI models until a competing alternative emerges. And that’s not to mention AI models from outside the US like Qwen3.5.
In a blog post, Google’s Gemini team claims that Gemini 3.1 Pro can tackle complex problem solving better than previous models. And they cite benchmark test results to support their claims, which should be viewed with a degree of skepticism. On the ARC-AGI-2 problem-solving test, Gemini 3.1 Pro scored 77.1 percent, compared to Gemini 3 Pro’s score of 31.1 percent and Gemini 3 Deep Think’s score of 45.1 percent.
According to Google’s graphs, Gemini 3.1 Pro outperforms competing commercial models such as Anthropic’s Opus 4.6 and Sonnet 4.6, and OpenAI’s GPT-5.2 and GPT-5.3-Codex on most of the benchmarks cited. However, Opus 4.6 maintains the highest scores for Last of Mankind (Full Set, Test + MM), SWE-Bench Verified, and τ²-bench. GPT-5.3-Codex also leads in SWE-Bench Pro (public) and Terminal-Bench 2.0 when evaluated using Codex’s own harness rather than the standard Terminus-2 agent harness.
“3.1 Pro is designed for tasks where simple answers aren’t enough, and employs advanced reasoning to help you with your toughest challenges,” said the Gemini team. “This increased intelligence lends itself to practical applications, whether you’re looking for clear visual explanations of complex topics, ways to integrate data into a single view, or ways to bring creative projects to life.”
To illustrate potential uses, the Gemini team pointed out how this model can create website-ready SVG animations that can translate the literary style of a novel into the design of a personal portfolio site.
In the company’s fourth quarter 2025 financial results announcement. [PDF]CEO Sundar Pichai said, “Our first-party models like Gemini currently process over 10 billion tokens per minute through direct use of our API by our customers, and the Gemini app has grown to over 750 million monthly active users.”
Google is making Gemini 3.1 Pro available through the Gemini API in Google AI Studio, Gemini CLI, Antigravity, and Android Studio. Business customers can access it via Vertex AI and Gemini Enterprise, and consumers can access it via the Gemini app and NotebookLM.
This model is also accessible from several Microsoft services, including GitHub Copilot, Visual Studio, and Visual Studio Code. ®
