VECTOR WIREAI INTELLIGENCE
⌘K
PKT
Refresh Models —Deals —Regulatory —Sources —

Questions this page answers

Which is better, Qwen2 VL 72B Instruct or Qwen3 VL Thinking (8B)?
Across 13 shared benchmarks, Qwen2 VL 72B Instruct scores higher on 5 and Qwen3 VL Thinking (8B) on 8. The widest gap is mathvision, where Qwen3 VL Thinking (8B) scores 62.7 against 25.9.

Qwen2 VL 72B Instruct vs Qwen3 VL Thinking (8B)

Across 13 shared benchmarks, Qwen2 VL 72B Instruct scores higher on 5 and Qwen3 VL Thinking (8B) on 8. The widest gap is mathvision, where Qwen3 VL Thinking (8B) scores 62.7 against 25.9.

AlibabavsAlibaba13 shared benchmarks5–8 head-to-head
BenchmarkQwen2 VL 72B InstructQwen3 VL Thinking (8B)
CC-OCR68.776.3
DocVQA_test96.595.3
InfoVQA_test84.586
mathvision25.962.7
MathVista70.581.4
MMMU64.573.5
MMMU (val) (Pass@1)64.574.1
MMMU-Pro46.260.4
MMStar68.375.3
MV-Bench73.669
OCRBench87782.8
RealWorldQA77.873.5
Video-MME77.871.8

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.