MedicalBenchmark
Meta: Llama 3.1 8B Instruct provider

Llama 3.1 8B Instruct

298

#298 of 319 models in the general ranking

Cumulative performance across 3 MIR exams

Net score

160.33 pts

Accuracy

39.8%

Correct / Incorrect

239 / 236

Total Cost

$0.06

Overall Performance

(vs. average)
Accuracy

39.8%

avg: 80.6%

Net score

160.33 pts

avg: 453.30 pts

Correct

239

avg: 483

Incorrect

236

avg: 90

Total Cost

$0.06

avg: $9.58

Average response time

25.0s

avg: 17.9s

Output Tokens

784K

avg: 1.3M

Reasoning Tokens

0

avg: 898K

Average confidence

79.1%

avg: 95.4%

Breakdown by Exam

MIR 2024
303
Correct
74
Incorrect
90
Accuracy
37.0%
Net score
44.00
MIR 2025
290
Correct
79
Incorrect
78
Accuracy
39.5%
Net score
53.00
MIR 2026
299
Correct
86
Incorrect
68
Accuracy
43.0%
Net score
63.33