Gemini 4 Carbon 登顶 AA 智能指数，59 分压过 Opus
博主称 Gemini 4 Carbon 在 Artificial Analysis Intelligence Index v5.0 拿到 59 分、超过 Opus 的 58 分，并提到 Arena 上 Argon 检查点被路由到 Carbon 后实测质量出乎意料。
🔴 Gemini 4 Carbon is now the best model on the Artificial Analysis Intelligence Index v5.0, even beating Opus with a score of 59 vs 58! This is such a huge achievement for Google, I've been testing the new Gemini 4 Argon checkpoint on Arena that was redirecting to Carbon and the quality has really surprised me. Just think about it, Google went almost 8 months without a major update after Gemini 3.1 Pro and now they've managed to release a model that's outperforming Anthropic and OpenAI in these benchmarks. I never expected such a big jump in performance, especially after such a long wait, and seeing Google back at the top of the rankings is something I really didn't see coming. I'm pretty much speechless!