New AI Model Rankings: Zhipu GLM-5.3-Max Debuts in Top 15, Kimi K3-Max Enters Top 10

Arena, an anonymous AI model ranking platform, released its latest weekly rankings. Zhipu AI's GLM-5.3-Max debuted at No. 13 in the overall list with an ELO score of 1487, while Moonshot's Kimi-K3-Max climbed to No. 10, entering the top 10 for the first time. Chinese models showed strong performance in several categories. In the overall list, Anthropic's Claude series took the top three spots, with Claude-Fable-5 leading at 1508 points. Alibaba's Qwen3.8-Max dropped to No. 19, and Qwen3.7-Max-Preview ranked 27th. In the coding list, GLM-5.3-Max debuted in the top 10, and Kimi-K3-Max stayed at No. 6. In the frontend development list, Kimi-K3-Max and Qwen3.8-Max ranked second and third, with GLM-5.3-Max entering at No. 8. In the AI image generation list, ByteDance's Seedream-5.0-Pro and Alibaba's Qwen-Image-3.0-Pro took 8th and 9th places. The rankings use an ELO system, where differences within the margin of error are considered ties, reflecting user preferences.