The Model Index
Up to four models

Gemini 2.5 Pro vs DeepSeek-R1

On the index (GPT-4 = 100), Gemini 2.5 Pro scores 141 and DeepSeek-R1 130. 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 Gemini 2.5 Pro
Presets
Gemini 2.5 Pro
141125–157
WeightsClosed
ReleasedMar 2025
Price in / out–
Context–
Core tests taken2 of 9
DeepSeek-R1
130114–146
WeightsOpen
ReleasedJan 2025
Price in / out–
Context–
Core tests taken2 of 9

Gemini 2.5 Pro leads DeepSeek-R1 by 11 points, ahead on 1 of 1 shared core tests.

Test by testthird-party results

Gemini 2.5 ProDeepSeek-R1
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)Fiction.LiveBenchFiction.LiveBenchSimpleBenchSimpleBenchMATHSMATH (level 5)MATH (level 5)CODINGAider PolyglotAider PolyglotAGENTSBALROGBALROG
Every result as a table
TestGemini 2.5 ProDeepSeek-R1
Reasoning & knowledge
GPQA Diamond (Epoch AI)83.8%71.7%
Fiction.LiveBench66.7%69.4%
SimpleBench51.6%30.9%
Maths
MATH (level 5)95.6%93.0%
Coding
Aider Polyglot72.9%56.9%
Agents
BALROG43.3%34.9%
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