Hey there,

Today’s AI models are already great at simple, everyday tasks. The real challenge comes when they need to perform complex reasoning and agentic tasks. Google’s latest state-of-the-art model Gemini 3.1 Pro was designed with this fact in mind. From increased reasoning capabilities to new SVG animations, let’s explore what the new Gemini model is offering. 

A Smarter and More Efficient Engine

The release of Gemini 3.1 Pro doesn’t bring a long list of brand new features but instead focuses on improving the core capabilities of the model. Let’s take a look at what’s changed:

Improved Reasoning: 

This is the headline improvement of Gemini 3.1 Pro. This new model is able to solve “abstract reasoning puzzles” at nearly double the rate of Gemini 3 Pro. This means that you can give larger, more complex issues to Gemini and get real solutions. 

More Control:

Gemini 3.1 Pro also gives us more control over how the model operates. The expanded thinking levels now include “Medium”. This allows us to better control the balance of speed, cost, and performance. These expanded thinking levels are mostly useful for developers using the API. 

Code-Based Animation: 

The improved reasoning allows Gemini to generate complex, website-ready animated SVGs from basic text prompts. These code-based animations offer a smaller file size and a more scalable option for web design. 

How to Access Gemini 3.1 Pro

• Users: Gemini 3.1 Pro is rolling out across the Gemini app, NotebookLM, and Google AI Studio. Currently it's limited to Google AI Pro and Ultra plans but will be coming to free users in the near future. 

• Developers: The model is available via the Gemini API in Google AI Studio, the Gemini CLI, and Google Antigravity. 

How It Compares to Other Models

When looking at new models, it's always important to look at how they compare to what's currently on the market. The benchmark comparison that we looked at earlier shows that Gemini 3.1 Pro brings improvements across the board. Most notably, the increases around agentic search and abstract reasoning makes Gemini 3.1 Pro a great choice for vibe coding. It's also important to look past the benchmarks shared by Google. In this case, we can look at LMArena. This platform allows users to blindly compare and rank models based on responses. Here we can see that Gemini 3.1 Pro is outperforming most models but not all. Opus 4.6 is consistently rated higher than Gemini 3.1 Pro across the board. 

My Thoughts: 

Google’s latest model seems to be following the general AI trend of improving complex and agentic reasoning. We discussed before that there have been an explosion of agentic AI experiences and now we’re getting models that are optimized for this new reality. Based on benchmarks and LMArena, it seems clear that Gemini 3.1 Pro can compete with other state-of-the-art models like Opus 4.6 and GPT-5.2 Thinking. I’m also looking forward to exploring 3.1 Pro’s ability to create animated SVG files. This seems like a great application of AI for both web design and the creation of professional looking slides and graphics. I wouldn’t let the “.1” in the name fool you. This new model offers better performance and great new capabilities that make it worth exploring. 

Of course the best way to judge this new model is to experience it yourselves. I highly recommend you check out Gemini 3.1 Pro’s coding abilities by vibe coding in Google AI Studio.