The Model Index
Up to four models

Llama 3-8B vs Llama 3.1-8B

On the index (GPT-4 = 100), Llama 3-8B scores 89 and Llama 3.1-8B 89. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to Llama 3-8B
Presets
Llama 3-8B
8971–108
WeightsOpen
ReleasedApr 2024
Price in / out–
Context–
Core tests taken1 of 9
Llama 3.1-8B
8970–107
WeightsOpen
ReleasedJul 2024
Price in / out$0.020 / $0.040
Context131K
Core tests taken1 of 9

Llama 3-8B and Llama 3.1-8B are too close to call (1 point apart, within the uncertainty), ahead on 0 of 1 shared core tests.

Test by testthird-party results

Llama 3-8BLlama 3.1-8B
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)MMLU (HELM runs)MMLU (HELM runs)MATHSAIME-style maths (OTIS mock)AIME-style maths (OTIS mock)MATH (level 5)MATH (level 5)AGENTSDTBenchDTBench
Every result as a table
TestLlama 3-8BLlama 3.1-8B
Reasoning & knowledge
GPQA Diamond (Epoch AI)26.1%27.0%
MMLU (HELM runs)68.8%56.1%
Maths
AIME-style maths (OTIS mock)1.9%1.7%
MATH (level 5)6.1%22.9%
Agents
DTBench43.9%50.9%

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG