The Model Index
Up to four models

DeepSeek-R1 vs GPT-4.5

On the index (GPT-4 = 100), DeepSeek-R1 scores 130 and GPT-4.5 125. 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-R1
Presets
DeepSeek-R1
130114–146
WeightsOpen
ReleasedJan 2025
Price in / out–
Context–
Core tests taken2 of 9
GPT-4.5
125109–141
WeightsClosed
ReleasedFeb 2025
Price in / out–
Context–
Core tests taken3 of 9

DeepSeek-R1 leads GPT-4.5 by 5 points, ahead on 1 of 1 shared core tests.

Test by testthird-party results

DeepSeek-R1GPT-4.5
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)Fiction.LiveBenchFiction.LiveBenchSimpleBenchSimpleBenchMATHSAIME-style maths (OTIS mock)AIME-style maths (OTIS mock)MATH (level 5)MATH (level 5)CODINGAider PolyglotAider PolyglotWeirdMLWeirdML
Every result as a table
TestDeepSeek-R1GPT-4.5
Reasoning & knowledge
GPQA Diamond (Epoch AI)71.7%68.7%
Fiction.LiveBench69.4%63.9%
SimpleBench30.9%34.5%
Maths
AIME-style maths (OTIS mock)53.3%37.8%
MATH (level 5)93.0%78.6%
Coding
Aider Polyglot56.9%44.9%
WeirdML36.5%39.4%
Task length · METR31 sec

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

This page as a chartDownload PNG