MedicalBenchmark
Meta: Llama 3 8B Instruct provider

Llama 3 8B Instruct

294

#294 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

176.66 pts

Accuracy

44.0%

Correct / Incorrect

264 / 262

Total Cost

$0.03

Overall Performance

(vs. average)
Accuracy

44.0%

avg: 80.6%

Net score

176.66 pts

avg: 453.30 pts

Correct

264

avg: 483

Incorrect

262

avg: 90

Total Cost

$0.03

avg: $9.58

Average response time

11.6s

avg: 17.9s

Output Tokens

390K

avg: 1.3M

Reasoning Tokens

0

avg: 898K

Average confidence

85.5%

avg: 95.4%

Breakdown by Exam

MIR 2024
288
Correct
93
Incorrect
83
Accuracy
46.5%
Net score
65.33
MIR 2025
300
Correct
72
Incorrect
101
Accuracy
36.0%
Net score
38.33
MIR 2026
291
Correct
99
Incorrect
78
Accuracy
49.5%
Net score
73.00