Compare · ModelsLive · 2 picked · head to head
Meta Llama 3 8B Instruct vs Meta Llama 3 8B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Meta Llama 3 8B Instruct wins on 6/11 benchmarks
Meta Llama 3 8B Instruct wins 6 of 11 shared benchmarks. Leads in language.
Category leads
general·Meta Llama 3 8Bknowledge·Meta Llama 3 8Blanguage·Meta Llama 3 8B Instructmath·Meta Llama 3 8Breasoning·Meta Llama 3 8B
Hype vs Reality
Attention vs performance
Meta Llama 3 8B Instruct
#113 by perf·no signal
Meta Llama 3 8B
#119 by perf·no signal
Vendor risk
Who is behind the model
Meta AI
$1.50T·Tier 1
Meta AI
$1.50T·Tier 1
Head to head
11 benchmarks · 2 models
Meta Llama 3 8B InstructMeta Llama 3 8B
BBH (HuggingFace)
Meta Llama 3 8B leads by +20.5
Meta Llama 3 8B Instruct
28.2
Meta Llama 3 8B
48.7
GPQA
Meta Llama 3 8B leads by +18.5
Meta Llama 3 8B Instruct
1.2
Meta Llama 3 8B
19.7
IFEval
Meta Llama 3 8B Instruct leads by +58.0
Meta Llama 3 8B Instruct
74.1
Meta Llama 3 8B
16.0
MATH Level 5
Meta Llama 3 8B leads by +9.9
Meta Llama 3 8B Instruct
8.7
Meta Llama 3 8B
18.6
MMLU-PRO
Meta Llama 3 8B leads by +11.6
Meta Llama 3 8B Instruct
29.6
Meta Llama 3 8B
41.2
MUSR
Meta Llama 3 8B leads by +14.4
Meta Llama 3 8B Instruct
1.6
Meta Llama 3 8B
16.0
JCommonsenseQA
Meta Llama 3 8B Instruct leads by +4.8
Meta Llama 3 8B Instruct
87.7
Meta Llama 3 8B
82.9
JMMLU
Meta Llama 3 8B Instruct leads by +2.0
Meta Llama 3 8B Instruct
46.7
Meta Llama 3 8B
44.7
JNLI
Meta Llama 3 8B Instruct leads by +0.2
Meta Llama 3 8B Instruct
61.1
Meta Llama 3 8B
60.9
JSQuAD
Meta Llama 3 8B Instruct leads by +0.6
Meta Llama 3 8B Instruct
89.5
Meta Llama 3 8B
88.9
LLM-JP · Overall
Meta Llama 3 8B Instruct leads by +0.8
Meta Llama 3 8B Instruct
49.6
Meta Llama 3 8B
48.9
Full benchmark table
| Benchmark | Meta Llama 3 8B Instruct | Meta Llama 3 8B |
|---|---|---|
BBH (HuggingFace) | 28.2 | 48.7 |
GPQA | 1.2 | 19.7 |
IFEval | 74.1 | 16.0 |
MATH Level 5 | 8.7 | 18.6 |
MMLU-PRO | 29.6 | 41.2 |
MUSR | 1.6 | 16.0 |
JCommonsenseQA | 87.7 | 82.9 |
JMMLU | 46.7 | 44.7 |
JNLI | 61.1 | 60.9 |
JSQuAD | 89.5 | 88.9 |
LLM-JP · Overall | 49.6 | 48.9 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| Meta Llama 3 8B Instruct | — | — | — | — |
| Meta Llama 3 8B | — | — | — | — |