Qwen3.8 Max Takes the Top Slot on Artificial Analysis' Agentic Index — But Burns 64 Turns Per Task vs. Qwen3.7's 14
Artificial Analysis / Hacker News·high signal
Qwen3.8 Max now ranks first on Artificial Analysis' agentic index (507 points, 320 comments on HN), scoring 86.1 on OSWorld-Verified ahead of GPT-5.6 Sol Max (83.2) and Fable 5 (85.0), at $2.00/M input and $6.00/M output. Its Intelligence Index score of 56 puts it level with Claude Opus 4.8 (max) and ahead of every model from Google, Meta and xAI. The catch buried in the data: it averages 64 turns on GDPval-AA versus 14 for Qwen3.7 Max, so the headline agentic win comes with a ~4.5x turn-count tax that a per-token price comparison hides entirely.