Intelligence Index v5：Gemini 4 Argon 得 50 分
Artificial Analysis 发布 Intelligence Index v5，Gemini 4 Argon（high）得 50 分，落后 Opus 5.5 的 58 与 GPT-6 Astra 的 57，闭源开源差距扩大到 12 分。
artificial analysis just dropped Intelligence Index v5, and open models got hit hardest. > Claude Opus 5.5 (max): 58 > GPT-6 Astra (max): 57 > GPT-6.1 Sol (max): 55 > Gemini 4 Argon (high): 50 > Kimi K3.1 Pro: 46 (weights not out yet) > Best open weights you can actually download: DeepSeek V4.1 Flash and GLM-5.3 at 32 the gap between closed and open was 3 points in v4.3. -now it's 12 what changed ? 60% of the index is now private test sets so there is less room to train on the test > Opus 5.5 max costs $10.04 per task > Sol max costs $1.27 and scores 3 points lower if you're paying per token, Sol is the deal and If you need the top score, you're paying 8x for it