Google has officially unveiled Gemini 3.1 Pro, marking the first entry in its 3.1 series with an aggressive pricing strategy targeting high-end competitors like Claude Opus 4.6. This iteration maintains the same cost structure as its predecessor while delivering benchmark scores that rival the industry's most capable models, effectively halving the cost relative to comparable high-intelligence models.

A standout feature of this release is its refined ability to generate Scalable Vector Graphics (SVG), a format for two-dimensional graphics that can be scaled without losing quality. In early testing, the model successfully interpreted complex prompts—such as a pelican riding a bicycle—incorporating intricate details like precise leg placement and internal code comments that describe specific plumage colors. This indicates a higher level of spatial and logical understanding when translating text into visual code.

The launch includes a specialized "custom tools" preview model, designed to offer superior performance for developers integrating the AI into specific software workflows. While early users reported significant latency and "high demand" errors during the initial rollout, the update represents a significant leap in the core intelligence driving Google's more specialized research tools.

The release follows the recent "Deep Think" update, suggesting that the 3.1 family is designed to handle more rigorous scientific and engineering tasks. By highlighting niche capabilities like SVG animation, Google is shifting its focus toward demonstrating concrete, functional improvements that differentiate its latest models from previous generations.

Google has officially unveiled Gemini 3.1 Pro, marking the first entry in its 3.1 series with an aggressive pricing strategy targeting high-end competitors like Claude Opus 4.6. This iteration maintains the same cost structure as its predecessor while delivering benchmark scores that rival the industry's most capable models, effectively halving the cost relative to comparable high-intelligence models.

A standout feature of this release is its refined ability to generate Scalable Vector Graphics (SVG), a format for two-dimensional graphics that can be scaled without losing quality. In early testing, the model successfully interpreted complex prompts—such as a pelican riding a bicycle—incorporating intricate details like precise leg placement and internal code comments that describe specific plumage colors. This indicates a higher level of spatial and logical understanding when translating text into visual code.

The launch includes a specialized "custom tools" preview model, designed to offer superior performance for developers integrating the AI into specific software workflows. While early users reported significant latency and "high demand" errors during the initial rollout, the update represents a significant leap in the core intelligence driving Google's more specialized research tools.

The release follows the recent "Deep Think" update, suggesting that the 3.1 family is designed to handle more rigorous scientific and engineering tasks. By highlighting niche capabilities like SVG animation, Google is shifting its focus toward demonstrating concrete, functional improvements that differentiate its latest models from previous generations.