Qwen 3 Max Instruct: benchmark results

Provider: Alibaba. Released 2025-09-24. Access: API.

Unified ELO 1748 ± 32, rank #242 of 2656 rated models, from 22 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AGI-Eval Community - Learning (Chinese)83.59Accuracy (%)80.4
AGI-Eval Community - Subject Reasoning (English)84.95Accuracy (%)77.5
AGI-Eval Community - Subject Reasoning86.07Accuracy (%)76.8
AGI-Eval Community - Subject Reasoning (Chinese)88.61Accuracy (%)76.8
AGI-Eval Community - Learning88.06Accuracy (%)75
AGI-Eval Community - Mathematical Reasoning79.84Accuracy (%)70
AGI-Eval Community - General Reasoning96.89Accuracy (%)68.9
AGI-Eval Community - Subject Knowledge85.28Accuracy (%)67.1
OTIS Mock AIME 2024-202573.33Score (%)63.5
AGI-Eval Community - Objective Accuracy (English)87.32Accuracy (%)61.6
AGI-Eval Community - Learning (English)90.99Accuracy (%)57.2
AGI-Eval Community - Interaction (Chinese)77.2Accuracy (%)55.8

Interactive version: theaggregate.ai/model?slug=qwen-3-max-instruct · How It Works · Data refreshed daily, snapshot 2026-09-19.