Alibaba (Qwen)
Qwen3-Max benchmark results
11 results on 6 benchmarks from 4 sources.
Also listed as: qwen3-max-2025-09-23 · Qwen3-Max-Instruct · Qwen3-Max-Thinking
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1423 | 125 | Overall, style control on; 9029 votes; 95% CI 1417–1430“qwen3-max-2025-09-23” | Official leaderboard | Oct 8, 2026 |
| GPQA Diamond | 72.6% | 151 | Epoch AI run, accuracy | Independent run | Oct 6, 2025 |
| MMLU-Pro | 85.7% | 31 | Overall accuracy; data source: Self-Reported“Qwen3-Max-Thinking” | Vendor-reported | Mar 11, 2026 |
| AIME (OTIS Mock AIME 2024–2025) | 73.3% | 120 | Epoch AI run, OTIS Mock AIME 2024–2025 | Independent run | Oct 6, 2025 |
| MATH Level 5 | 97.1% | 7 | Epoch AI run, accuracy | Independent run | Oct 9, 2025 |
| τ²-bench | 69.0% | 11 | airline domain, pass^1; submitted by Qwen; vv0.1.3“Qwen3-Max-Thinking” | Vendor-reported | Jan 23, 2026 |
| τ²-bench | 59.5% | 17 | airline domain, pass^1; submitted by Alibaba; v0.1.3 | Official leaderboard | Sep 23, 2024 |
| τ²-bench | 79.4% | 13 | retail domain, pass^1; submitted by Qwen; vv0.1.3“Qwen3-Max-Thinking” | Vendor-reported | Jan 23, 2026 |
| τ²-bench | 72.2% | 22 | retail domain, pass^1; submitted by Alibaba; v0.1.3 | Official leaderboard | Sep 23, 2024 |
| τ²-bench | 98.2% | 1 | telecom domain, pass^1; submitted by Qwen; vv0.1.3“Qwen3-Max-Thinking” | Vendor-reported | Jan 23, 2026 |
| τ²-bench | 84.2% | 15 | telecom domain, pass^1; submitted by Alibaba; v0.1.3 | Official leaderboard | Sep 23, 2024 |
FAQ
Where do these Qwen3-Max scores come from?
From Epoch AI Benchmarking Hub, LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), and τ²-bench leaderboard submissions. Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.