Llama 3.1
LatestMeta
•
Open Source•
8B# 50
Released
Jul 23, 2024
# 9
Knowledge Cutoff
Dec 23
# 7
Context Length
128K
Benchmarks
# 63
Code LiveBench
19.7%
# 48
Code LMArena
1174
# 61
GPQA Diamond
27.0%
# 52
Reason LiveBench
15.3%
# 52
ELO LMArena
1171
# 50
Math LiveBench
16.6%
# 34
MATH
64.0%
# 32
Human Eval
72.6%
# 21
Human Eval+
62.8%
# 29
Aider Old
37.6%
# 49
MMLU Pro
44.3%
# 33
MMLU
71.0%
# 25
Halluc. Hughes
5.4%
Pricing
# 16
Input Cost /M
$0.3
# 15
Output Cost /M
$0.6