Qwen4 Max lands on PostTrainBench at 36.7%, edging below GLM 5.3 Flash

maksym_andr · x · 2026-11-20

Qwen's biggest new flagship, Qwen4 Max, has been added to the PostTrainBench leaderboard (v1.2), scoring 36.7% ±0.9 — slightly below GLM 5.3 Flash at 36.8% and on par with GLM 5.3 and Opus 5 (both 35.7%). It trails far behind the leaderboard's top entry, Official Instruct at 48.4%, and Fable 5.2 at 44.6%. Tested with Qwen Code at xHigh reasoning, averaging a 9:16 runtime.

Related event: Qwen4 Max Scores 36.7% on PostTrainBench, On Par With GLM 5.3 Flash(2 posts)→

Original post →

More from Models

Models channel →