Qwen3-Max-Thinking
Alibaba·
LLMsreasoningcloud + local
Overview
Trillion-parameter reasoning model with adaptive tool use and test-time scaling.
Capabilities and innovations
Advanced reasoningAdaptive tool useTest-time scalingCode interpretationWeb search integrationVideo understandingReasoning ModeAdaptive Tool Invocation1T+ MoE ArchitectureDynamic Compute Allocation
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 87.4% | unsourced |
| Humanity’s Last Exam (no tools) | 36.5% | third party HLE 36.5 no-tools (with test-time scaling); base no-tools 34.1. Alibaba also reports 49.8 with web-search tools and 58.3 with tools + test-time scaling — not comparable to the no-tools value stored here. Corrected from a previously stored with-tools value (58.3). |
| SWE-bench Verified | 75.3% | unsourced |
| MMLU-Pro | 84.98% | unsourced |
- SWE-bench Pro: not reported by the vendor
- MMLU: not reported by the vendor
- Terminal-Bench 2.x: not reported by the vendor
- MMMU-Pro: not reported by the vendor
Links
More from Alibaba
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.