MedicalBenchmark
StepFun: Step3 provider

Step3

189

#189 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

485.66 pts

Accuracy

85.7%

Correct / Incorrect

514 / 85

Total Cost

$1.82

Overall Performance

(vs. average)
Accuracy

85.7%

avg: 80.6%

Net score

485.66 pts

avg: 453.30 pts

Correct

514

avg: 483

Incorrect

85

avg: 90

Total Cost

$1.82

avg: $9.58

Average response time

51.1s

avg: 17.9s

Output Tokens

1.2M

avg: 1.3M

Reasoning Tokens

674K

avg: 898K

Average confidence

99.6%

avg: 95.4%

Breakdown by Exam

MIR 2024
199
Correct
172
Incorrect
28
Accuracy
86.0%
Net score
162.66
MIR 2025
163
Correct
170
Incorrect
29
Accuracy
85.0%
Net score
160.33
MIR 2026
199
Correct
172
Incorrect
28
Accuracy
86.0%
Net score
162.66