Google has unveiled Gemini 3.1 Pro, a significant upgrade to its artificial intelligence model lineup, stating that the new system more than doubles the reasoning performance of its predecessor.
Released in preview, Gemini 3.1 Pro is designed to move artificial intelligence beyond simple text generation toward what the company describes as agentic workflows. These are autonomous systems capable of solving complex, multi step problems across domains such as software engineering, finance, and research.
According to Google, the model builds on the architecture introduced in the Gemini 3 series launched last November. Internal benchmarks show that Gemini 3.1 Pro achieved a score of 77.1 percent on ARC AGI 2, a rigorous evaluation that measures a model’s ability to solve entirely new logic patterns it has not encountered during training.
The score represents more than twice the performance of the 3.0 Pro version and narrows the gap with last week’s specialised “Deep Think” mode, which focuses on enhanced reasoning tasks.
Focus On Complex Task Execution
The upgrade reflects a broader industry shift toward AI systems capable of structured, multi-stage task execution. Rather than responding to isolated prompts, Gemini 3.1 Pro is positioned to plan, execute, and refine outputs across extended workflows.
In software engineering, this could translate into analysing large codebases, identifying bugs, and generating structured solutions. In finance, the system may assist with layered data interpretation and structured reporting. In research, it could support literature synthesis and multi-step analytical tasks.
By strengthening reasoning performance, Google aims to improve reliability and reduce errors in high-stakes use cases where accuracy and logical consistency are critical.
Direct Generation Of Animated SVGs
A standout feature of Gemini 3.1 Pro is its ability to generate website-ready, animated Scalable Vector Graphics directly from a text prompt. Users can describe visual elements and animations in natural language, with the system producing structured SVG outputs suitable for web deployment.
This capability extends the model’s utility beyond text and code generation into dynamic visual content creation, offering potential value for designers, developers, and digital marketing teams.
Competitive AI Landscape
The release comes amid intensifying competition in the generative AI sector, with major technology companies racing to enhance reasoning depth and deploy more autonomous systems. By emphasising benchmark gains and agentic functionality, Google is signalling that the next phase of AI development lies in structured problem solving rather than conversational output alone.
As Gemini 3.1 Pro enters preview, developers and enterprise users will evaluate whether the benchmark improvements translate into measurable gains in real-world applications. If performance aligns with internal testing results, the update could mark a meaningful step toward AI systems capable of executing complex tasks with greater independence and precision.
