Qwen 3.6 Max Preview — benchmark results

Alibaba's preview of the Qwen 3.6 Max flagship tier. Provider: Alibaba. Released 2026-04-20. Access: API.

Unified ELO 1791 ± 12, rank #146 of 1776 rated models, from 121 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
OpenCompass Language - Instruction Following80Score (%)100
PawBench - QwenPaw78.35Overall Score (%)100
AI for Education Pedagogy - Science94.54Accuracy (%)99.1
AI for Education Pedagogy - Primary94.37Accuracy (%)97.7
Wolfram LLM Benchmarking Project68.7Correct Functionality (%)97.6
AA Omniscience - Software Engineering (SWE) - Rust82Accuracy (%)97.5
AI for Education Pedagogy89.66Accuracy (%)96.8
AA TAU-2 Bench95.91Accuracy (%)96.2
AI for Education Pedagogy - Maths91.27Accuracy (%)96.1
AI for Education Pedagogy - Secondary88.99Accuracy (%)96.1
AA IFBench76.6Accuracy (%)95.8
OpenCompass LLM - Language79.5Score (%)95.5

Interactive version: theaggregate.ai/model?slug=qwen-3-6-max-preview · How the rankings work · Data refreshed daily, snapshot 2026-07-22.