Neura Intelligence Index

AI Model Benchmarks

Back to Leaderboard

DeepSeek-V3.2-Speciale vs LongCat-Flash-Thinking

DeepSeek-V3.2-Speciale

DeepSeek 🇨🇳

59.2

#153—

M

LongCat-Flash-Thinking

Meituan 🇨🇳

59.0

#156—

Category Comparison

Benchmark-by-Benchmark

math

Benchmark	DeepSeek-V3.2-Speciale	LongCat-Flash-Thinking
AIME 2025	96	90.6

reasoning

Benchmark	DeepSeek-V3.2-Speciale	LongCat-Flash-Thinking
Humanity's Last Exam	30.6	—
GPQA Diamond	—	81.5

coding

Benchmark	DeepSeek-V3.2-Speciale	LongCat-Flash-Thinking
SWE-bench Verified	73.1	59.4

agent

Benchmark	DeepSeek-V3.2-Speciale	LongCat-Flash-Thinking
Toolathlon	35.2	—

Category Winners

coding

DeepSeek-V3.2-Speciale

67.3

math

DeepSeek-V3.2-Speciale

79.4

reasoning

M

LongCat-Flash-Thinking

74.4

agent

DeepSeek-V3.2-Speciale

11.2

DeepSeek-V3.2-Speciale vs LongCat-Flash-Thinking — Benchmark Comparison | Neura Market