MedicalBenchmark
Meta: Llama 3.1 70B Instruct provider

Llama 3.1 70B Instruct

206

#206 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

467.33 pts

Accuracy

82.8%

Correct / Incorrect

497 / 89

Total Cost

$0.26

Overall Performance

(vs. average)
Accuracy

82.8%

avg: 80.6%

Net score

467.33 pts

avg: 453.30 pts

Correct

497

avg: 483

Incorrect

89

avg: 90

Total Cost

$0.26

avg: $9.58

Average response time

10.0s

avg: 17.9s

Output Tokens

271K

avg: 1.3M

Reasoning Tokens

0

avg: 898K

Average confidence

96.5%

avg: 95.4%

Breakdown by Exam

MIR 2024
203
Correct
170
Incorrect
28
Accuracy
85.0%
Net score
160.66
MIR 2025
204
Correct
161
Incorrect
33
Accuracy
80.5%
Net score
150.00
MIR 2026
214
Correct
166
Incorrect
28
Accuracy
83.0%
Net score
156.66