Neura Intelligence Index

AI Model Benchmarks

Back to Leaderboard

GPT-4 Turbo vs Llama 3.3 70B

OpenAI 🇺🇸

59.7

#114—

Meta 🇺🇸

59.3

#118—

Category Comparison

Benchmark-by-Benchmark

reasoning

Benchmark	GPT-4 Turbo	Llama 3.3 70B
GPQA Diamond	49.3	49.5
BigBench-Hard	85.1	82.9
ARC-Challenge	96.3	95.2

coding

Benchmark	GPT-4 Turbo	Llama 3.3 70B
HumanEval	87.6	88.4

math

Benchmark	GPT-4 Turbo	Llama 3.3 70B
GSM8K	94.8	95.8
MATH	72.2	77

language

Benchmark	GPT-4 Turbo	Llama 3.3 70B
HellaSwag	95.3	94.8
WinoGrande	87.5	86.1

safety

Benchmark	GPT-4 Turbo	Llama 3.3 70B
TruthfulQA	70.1	62.8

general

Benchmark	GPT-4 Turbo	Llama 3.3 70B
IFEval	82.1	85.3
MMLU-Pro	64.1	68.9
Chatbot Arena Elo	1256	1247

Category Winners

coding

Llama 3.3 70B

66.8

math

Llama 3.3 70B

71.9

reasoning

GPT-4 Turbo

45.5

general

Llama 3.3 70B

60.5

language

GPT-4 Turbo

85.2

safety

GPT-4 Turbo

84.8

GPT-4 Turbo vs Llama 3.3 70B — Benchmark Comparison | Neura Market