Rankings / Alibaba
Qwen3 Max
released 2025-09-23
BenchAtlas Index
as of 2026-07-14
base configuration
rank #269
9 families · 5 categories · high
base configuration
rank #274
8 families · 3 categories · medium
Benchmark evidence
19 results
agentic coding
| τ²-Bench | 74.3 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | 76.7 % | independent | Artificial Analysis | |
| SciCode | 38.3 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | 24.0 points | independent | Artificial Analysis | |
| AA Math Index | 80.7 points | independent | Artificial Analysis |
knowledge science
| GPQA Diamond implementation=vals-ai | 79.5 % | independent | Vals AI | |
| GPQA Diamond implementation=artificial-analysis | 76.4 % | independent | Artificial Analysis | |
| Humanity's Last Exam implementation=artificial-analysis | 11.1 % | independent | Artificial Analysis | |
| MMLU-Pro implementation=vals-ai | 84.4 % | independent | Vals AI | |
| MMLU-Pro implementation=artificial-analysis | 84.1 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | 46.7 % | independent | Artificial Analysis | |
| IFBench | 44.1 % | independent | Artificial Analysis |
professional
| CaseLaw | 47.5 % | independent | Vals AI | |
| CorpFin | 55.9 % | independent | Vals AI | |
| LegalBench | 81.9 % | independent | Vals AI | |
| MedQA | 87.4 % | independent | Vals AI | |
| TaxEval | 73.5 % | independent | Vals AI |
reasoning math
| AIME implementation=vals-ai | 81.0 % | independent | Vals AI | |
| AIME year=2025 · implementation=artificial-analysis | 80.7 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Qwen3 Max | Terminal-Bench Hard | 20.4 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.