Neura Intelligence Index

AI Model Benchmarks

Llama 3.3 70B vs Ministral 3 (14B Reasoning 2512) — Benchmark Comparison | Neura Market

Back to Leaderboard

Llama 3.3 70B vs Ministral 3 (14B Reasoning 2512)

Meta 🇺🇸

59.2

#139—

Ministral 3 (14B Reasoning 2512)

Mistral AI 🇫🇷

58.5

#142—

Category Comparison

Benchmark-by-Benchmark

language

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
HellaSwag	94.8	—
WinoGrande	86.1	—

math

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
MATH	77	—
GSM8K	95.8	—
AIME 2025	—	85

coding

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
HumanEval	88.4	—

reasoning

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
BigBench-Hard	82.9	—
GPQA Diamond	49.5	71.2
ARC-Challenge	95.2	—

safety

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
TruthfulQA	62.8	—

general

Benchmark	Llama 3.3 70B	Ministral 3 (14B Reasoning 2512)
MMLU-Pro	68.9	—
IFEval	85.3	—
Chatbot Arena Elo	1247	—

Category Winners

coding

Llama 3.3 70B

66.8

math

Llama 3.3 70B

71.9

reasoning

Ministral 3 (14B Reasoning 2512)

56.7

general

Llama 3.3 70B

60.5

language

Llama 3.3 70B

77.3

safety

Llama 3.3 70B

43.7