MedicalBenchmark
Qwen: Qwen3 VL 8B Thinking provider

Qwen3 VL 8B Thinking

205

#205 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

467.66 pts

Accuracy

83.3%

Correct / Incorrect

500 / 97

Total Cost

$4.00

Overall Performance

(vs. average)
Accuracy

83.3%

avg: 80.6%

Net score

467.66 pts

avg: 453.30 pts

Correct

500

avg: 483

Incorrect

97

avg: 90

Total Cost

$4.00

avg: $9.58

Average response time

39.2s

avg: 17.9s

Output Tokens

1.9M

avg: 1.3M

Reasoning Tokens

1.6M

avg: 898K

Average confidence

99.2%

avg: 95.4%

Breakdown by Exam

MIR 2024
208
Correct
168
Incorrect
30
Accuracy
84.0%
Net score
158.00
MIR 2025
177
Correct
168
Incorrect
32
Accuracy
84.0%
Net score
157.33
MIR 2026
232
Correct
164
Incorrect
35
Accuracy
82.0%
Net score
152.33