MedicalBenchmark
Qwen: Qwen3 VL 235B A22B Thinking provider

Qwen3 VL 235B A22B Thinking

79

#79 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

549.33 pts

Accuracy

93.7%

Correct / Incorrect

562 / 38

Total Cost

$3.56

Overall Performance

(vs. average)
Accuracy

93.7%

avg: 80.6%

Net score

549.33 pts

avg: 453.30 pts

Correct

562

avg: 483

Incorrect

38

avg: 90

Total Cost

$3.56

avg: $9.58

Average response time

37.5s

avg: 17.9s

Output Tokens

844K

avg: 1.3M

Reasoning Tokens

566K

avg: 898K

Average confidence

99.7%

avg: 95.4%

Breakdown by Exam

MIR 2024
88
Correct
189
Incorrect
11
Accuracy
94.5%
Net score
185.33
MIR 2025
80
Correct
182
Incorrect
18
Accuracy
91.0%
Net score
176.00
MIR 2026
71
Correct
191
Incorrect
9
Accuracy
95.5%
Net score
188.00