The Model Index
Up to four models

DeepSeek-V3.2-Speciale vs o4-mini

On the index (GPT-4 = 100), DeepSeek-V3.2-Speciale scores 143 and o4-mini 141. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to DeepSeek-V3.2-Speciale
Presets
DeepSeek-V3.2-Speciale
143125–162
WeightsOpen
ReleasedDec 2025
Price in / out–
Context–
Core tests taken0 of 9
o4-mini
141126–157
WeightsClosed
ReleasedApr 2025
Price in / out–
Context–
Core tests taken5 of 9

DeepSeek-V3.2-Speciale and o4-mini are too close to call (2 points apart, within the uncertainty).

Test by testthird-party results

DeepSeek-V3.2-Specialeo4-mini
0%25%50%75%100%REASONING & KNOWLEDGESimpleBenchSimpleBenchCODINGWeirdMLWeirdML
Every result as a table
TestDeepSeek-V3.2-Specialeo4-mini
Reasoning & knowledge
SimpleBench52.6%38.7%
Coding
WeirdML46.7%52.6%
Task length · METR38 sec

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG