MedicalBenchmark
Meta: Llama 3 70B Instruct provider

Llama 3 70B Instruct

237

#237 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

426.66 pts

Accuracy

78.0%

Correct / Incorrect

468 / 124

Total Cost

$0.31

Overall Performance

(vs. average)
Accuracy

78.0%

avg: 80.6%

Net score

426.66 pts

avg: 453.30 pts

Correct

468

avg: 483

Incorrect

124

avg: 90

Total Cost

$0.31

avg: $9.58

Average response time

13.8s

avg: 17.9s

Output Tokens

217K

avg: 1.3M

Reasoning Tokens

0

avg: 898K

Average confidence

97.9%

avg: 95.4%

Breakdown by Exam

MIR 2024
239
Correct
158
Incorrect
40
Accuracy
79.0%
Net score
144.66
MIR 2025
243
Correct
147
Incorrect
51
Accuracy
73.5%
Net score
130.00
MIR 2026
233
Correct
163
Incorrect
33
Accuracy
81.5%
Net score
152.00