Neura Intelligence Index

AI Model Benchmarks

Back to Leaderboard

DeepSeek-V3.2-Exp vs Mistral Medium 3.5

DeepSeek-V3.2-Exp

DeepSeek 🇨🇳

63.3

#99—

Mistral Medium 3.5

Mistral AI 🇫🇷

64.9

#96—

Category Comparison

Benchmark-by-Benchmark

reasoning

Benchmark	DeepSeek-V3.2-Exp	Mistral Medium 3.5
GPQA Diamond	79.9	—
Humanity's Last Exam	19.8	—

math

Benchmark	DeepSeek-V3.2-Exp	Mistral Medium 3.5
AIME 2025	89.3	86.3

coding

Benchmark	DeepSeek-V3.2-Exp	Mistral Medium 3.5
SWE-bench Verified	67.8	77.6
TerminalBench	37.7	—

general

Benchmark	DeepSeek-V3.2-Exp	Mistral Medium 3.5
SimpleQA	97.1	—

agent

Benchmark	DeepSeek-V3.2-Exp	Mistral Medium 3.5
BrowseComp	40.1	48.6

Category Winners

coding

Mistral Medium 3.5

77.0

math

DeepSeek-V3.2-Exp

68.6

reasoning

DeepSeek-V3.2-Exp

50.6

general

DeepSeek-V3.2-Exp

98.8

agent

Mistral Medium 3.5

21.0

DeepSeek-V3.2-Exp vs Mistral Medium 3.5 — Benchmark Comparison | Neura Market