Google's Gemini 4 Pro model has stepped onto the large model battleground, Arena, under the guise of the anonymous identifier "gemini-3.8-flash." Touted as Google's next-generation flagship model, it is the culmination of six months of intensive research and development. In benchmark tests spanning multiple domains—such as coding, agent capabilities, and reasoning—Gemini 4 Pro has decisively outperformed both GPT-6 Astra and Claude Fable 5.1, all while offering a more competitive pricing structure. Additionally, the model boasts a substantial 10 million input limit, permanent cross-session memory, and seamless internet connectivity without the need for an API.

In real-world tests, Gemini 4 Pro has showcased remarkable prowess in areas like UI web design, 3D+SVG generation, and game development, achieving high completion rates and delivering outstanding results. Google's strategic investment in RSI (Recursive Self-Improvement) technology is widely regarded as the linchpin behind this performance surge, with relevant research findings also being made publicly available.

Amidst the fierce competition among AI behemoths, Gemini 4 Pro faces the critical task of proving that the Gemini series can maintain its leading edge. This is pivotal for Google to uphold its market value. It's worth noting that the previous iteration, Gemini 3.5 Pro, was discontinued due to its relatively limited evolution compared to the Flash version.