From：Internet Info Agency 2026-08-10 17:32:00
On August 10, the Arena platform released its blind evaluation rankings for large AI models in Week 32 of 2026. Alibaba's Qwen3.8-Max ranked 6th on the overall leaderboard with an ELO score of 1,497, while Meta's Muse-Spark-1.2 (xHigh) secured 4th place overall. On the code-specific leaderboard, Anthropic's model took first place, followed by Moonshot AI's Kimi-K3-Max and Alibaba's Qwen3.8-Max at 6th and 8th, respectively. In frontend development, Alibaba's Qwen3.8-Max ranked 4th with an ELO score of 1,667. On the AI agent leaderboard, Moonshot AI's KimiK3 (Max) ranked 5th, achieving a net improvement of 10.08% and a task confirmation success rate of 14.97%. In the AI image generation category, Alibaba's Qwen-Image-3.0-Pro and ByteDance's Seedream-5.0-Pro ranked 7th and 8th, respectively. On the video generation leaderboard, MiniMax's Minimax-H3 debuted at 4th place with an ELO score of 1,455.
Zeekr 7X Overheats at Ningbo Charging Station; Involved Vehicle Had Unrepaired Collision Damage
Mercedes-AMG Unveils Teaser of New All-Electric SUV with Triple-Motor System
Mitsubishi’s ASX VR-e EV, Based on Foxtron Bria, Heads to Australia
Beijing Hyundai Ioniq V Opens Pre-sales on August 21, Launches in September
BYD Seal 06 2027 Model Launches August 11, Starting at RMB 105,000
XPeng G9L Unveils Five New Colors, Reveals Dimensions and Cabin Features
China's Renewable Hydrogen Production Exceeds 1.4 Million Tons, with 620 Refueling Stations
