Neura Intelligence Index

AI Model Benchmarks

Back to Leaderboard

DeepSeek R1 Distill Llama 70B vs MiniMax M2

DeepSeek R1 Distill Llama 70B

DeepSeek 🇨🇳

45.1

#194—

MiniMax 🇨🇳

45.7

#191—

Category Comparison

Benchmark-by-Benchmark

reasoning

Benchmark	DeepSeek R1 Distill Llama 70B	MiniMax M2
GPQA Diamond	65.2	78
Humanity's Last Exam	—	12.5

math

Benchmark	DeepSeek R1 Distill Llama 70B	MiniMax M2
AIME 2025	—	78

coding

Benchmark	DeepSeek R1 Distill Llama 70B	MiniMax M2
SWE-bench Verified	—	69.4
TerminalBench	—	46.3
SciCode	—	36

agent

Benchmark	DeepSeek R1 Distill Llama 70B	MiniMax M2
BrowseComp	—	44

Category Winners

coding

MiniMax M2

55.6

math

MiniMax M2

47.8

reasoning

DeepSeek R1 Distill Llama 70B

45.1

agent

MiniMax M2

15.4

DeepSeek R1 Distill Llama 70B vs MiniMax M2 — Benchmark Comparison | Neura Market