Rankings / Alibaba (Qwen)
Qwen3 7 Max
released 2026-05-19
BenchAtlas Index
as of 2026-07-14
base configuration
rank #65
6 families · 3 categories · medium
base configuration
rank #188
12 families · 5 categories · high
Benchmark evidence
22 results
agentic coding
| Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25 | 43.6 % | independent | LiveBench | |
| τ²-Bench | 94.7 % | independent | Artificial Analysis | |
| τ²-Bench subset=banking | 10.9 % | independent | Artificial Analysis |
coding
| Release 2026-06-25 tasks_counted=2 · livebench_version=2026-06-25 | 74.2 % | independent | LiveBench | |
| SciCode | 48.8 % | independent | Artificial Analysis |
data analysis
| Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25 | 71.8 % | independent | LiveBench |
external indices
| AA Coding Index | 66.0 points | independent | Artificial Analysis | |
| Intelligence Index v4.1 | 46.0 points | independent | Artificial Analysis | |
| Vals Index | 57.5 points | independent | Vals AI |
knowledge science
| GPQA Diamond implementation=artificial-analysis | 92.3 % | independent | Artificial Analysis | |
| GPQA Diamond implementation=vals-ai | 90.2 % | independent | Vals AI | |
| Humanity's Last Exam implementation=artificial-analysis | 38.1 % | independent | Artificial Analysis | |
| MMLU-Pro implementation=vals-ai | 89.3 % | independent | Vals AI |
language
| Release 2026-06-25 tasks_counted=3 · livebench_version=2026-06-25 | 79.7 % | independent | LiveBench |
long context instruction
| AA-LCR | 69.0 % | independent | Artificial Analysis | |
| IFBench | 80.5 % | independent | Artificial Analysis | |
| Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25 | 74.0 % | independent | LiveBench |
professional
| CorpFin | 63.7 % | independent | Vals AI | |
| LegalBench | 84.9 % | independent | Vals AI | |
| TaxEval | 75.3 % | independent | Vals AI |
reasoning math
| Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25 | 85.3 % | independent | LiveBench | |
| Release 2026-06-25 tasks_counted=4 · livebench_version=2026-06-25 | 83.3 % | independent | LiveBench |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Qwen3 7 Max | Terminal-Bench 2.1 | 74.5 % | independent | Artificial Analysis |
| agent + model Artificial Analysis harness + Qwen3 7 Max | Terminal-Bench Hard | 50.8 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.