MedicalBenchmark
OpenAI: GPT-3.5 Turbo (older v0613) provider

GPT-3.5 Turbo (older v0613)

114

#114 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

533.66 pts

Accuracy

91.5%

Correct / Incorrect

549 / 46

Total Cost

$0.88

Overall Performance

(vs. average)
Accuracy

91.5%

avg: 80.6%

Net score

533.66 pts

avg: 453.30 pts

Correct

549

avg: 483

Incorrect

46

avg: 90

Total Cost

$0.88

avg: $9.58

Average response time

7.5s

avg: 17.9s

Output Tokens

310K

avg: 1.3M

Reasoning Tokens

0

avg: 898K

Average confidence

99.0%

avg: 95.4%

Breakdown by Exam

MIR 2024
84
Correct
189
Incorrect
10
Accuracy
94.5%
Net score
185.66
MIR 2025
142
Correct
173
Incorrect
24
Accuracy
86.5%
Net score
165.00
MIR 2026
105
Correct
187
Incorrect
12
Accuracy
93.5%
Net score
183.00