TLDR
-
Gemini 3.1 Pro debuts, bringing stronger reasoning for complex workflows
-
ARC-AGI-2 hits 77.1%, as Google claims major gains in logic performance
-
Developers get preview access via Gemini API, AI Studio, and Android Studio
-
Vertex AI and Gemini Enterprise extend Gemini 3.1 Pro to business teams
-
Pricing stays at $2 per 1M input tokens, despite higher model capability
Alphabet Inc. (GOOG) traded at $304.64, up 0.23%, as Google introduced Gemini 3.1 Pro across its platforms. The company positioned the model as its most powerful AI system to date. The launch strengthens Google’s push to lead advanced reasoning in science, research, and engineering.
Stronger Core Reasoning Drives Performance Gains
Google released Gemini 3.1 Pro as an upgraded version of its Gemini 3 series. The company designed the model to handle complex tasks that require deeper reasoning. As a result, it targets workflows that demand planning, synthesis, and structured thinking.
The model achieved a verified 77.1% score on the ARC-AGI-2 benchmark. This result more than doubled the reasoning performance of Gemini 3 Pro. Consequently, Google demonstrated measurable progress in solving new logic patterns.
Internal benchmarks showed competitive results across coding and scientific domains. The model scored 94.3% on GPQA Diamond and reached an Elo of 2887 on LiveCodeBench Pro. Google positioned Gemini 3.1 Pro as a stronger baseline for long-horizon tasks.
Expanded Access Across Consumer and Enterprise Platforms
Google rolled out Gemini 3.1 Pro in preview for developers through the Gemini API. The company made it available in Google AI Studio, Gemini CLI, Google Antigravity, and Android Studio. At the same time, enterprises gained access through Vertex AI and Gemini Enterprise.
Consumers can now use the upgraded model through the Gemini app and NotebookLM. However, Google limited higher usage tiers to Google AI Pro and Ultra subscribers. This structure supports premium access while maintaining broader ecosystem integration.
Google also maintained its existing pricing structure for API users. Input costs remain at $2 per one million tokens for prompts up to 200,000 tokens. Therefore, developers receive higher reasoning performance without added pricing pressure.
Applied Intelligence Signals Competitive Shift
Google emphasized applied intelligence rather than chat-based interaction. The model can generate animated SVG graphics directly from text prompts. Because it produces code-based outputs, the files remain scalable and lightweight.
The system also demonstrated complex system synthesis capabilities. In one application, it configured telemetry data to build a live aerospace dashboard. It supported interactive 3D coding and creative web design outputs.
Industry partners reported measurable efficiency gains during early testing. JetBrains recorded a 15% quality improvement compared to prior versions. As competition intensifies, Google aims to lead through benchmark performance and enterprise reliability.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
