The Model Index
Up to four models

GPT-5 vs GPT-4.1

On the index (GPT-4 = 100), GPT-5 scores 149 and GPT-4.1 126. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

GPT-5GPT-4.1
Next to GPT-5
Presets
GPT-5
149133–164
WeightsClosed
ReleasedAug 2025
Price in / out$1.25 / $10
Context400K
Core tests taken6 of 9
GPT-4.1
126110–142
WeightsClosed
ReleasedApr 2025
Price in / out$2.0 / $8.0
Context1.05M
Core tests taken5 of 9

GPT-5 leads GPT-4.1 by 23 points, ahead on 5 of 5 shared core tests, at a similar price.

Test by testthird-party results

GPT-5GPT-4.1
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)Humanity's Last ExamHumanity's Last ExamGPQA Diamond (Vals AI)GPQA Diamond (Vals AI)MMLU ProMMLU ProMMMU ProMMMU ProFiction.LiveBenchFiction.LiveBenchGeoBenchGeoBenchSimpleBenchSimpleBenchSimpleQA VerifiedSimpleQA VerifiedMATHSFrontierMath (tiers 1-3)FrontierMath (tiers 1-3)AIME-style maths (OTIS mock)AIME-style maths (OTIS mock)MATH (level 5)MATH (level 5)MATH 500MATH 500MGSMMGSMChess puzzlesChess puzzlesFrontierMath (2025 set)FrontierMath (2025 set)FrontierMath tier 4 (2025 set)FrontierMath tier 4 (2025 set)CODINGSWE-bench VerifiedSWE-bench VerifiedAider PolyglotAider PolyglotLiveCodeBenchLiveCodeBenchWeirdMLWeirdMLAGENTSDTBenchDTBenchLMCALMCAPROFESSIONAL WORKCorpFin v2CorpFin v2LegalBenchLegalBenchMortgageTaxMortgageTaxTaxEval v2TaxEval v2NOVEL PROBLEMSARC-AGI-2ARC-AGI-2ARC-AGI-1ARC-AGI-1
Every result as a table
TestGPT-5GPT-4.1
Reasoning & knowledge
GPQA Diamond (Epoch AI)86.2%66.9%
Humanity's Last Exam25.3%5.4%
GPQA Diamond (Vals AI)85.6%65.4%
MMLU Pro86.5%80.5%
MMMU Pro81.5%72.4%
Fiction.LiveBench97.2%63.9%
GeoBench81.0%72.0%
SimpleBench56.7%27.0%
SimpleQA Verified50.1%31.1%
Maths
FrontierMath (tiers 1-3)55.4%6.0%
AIME-style maths (OTIS mock)91.4%38.3%
MATH (level 5)98.1%83.0%
MATH 50096.0%87.2%
MGSM92.8%87.7%
Chess puzzles36.0%11.0%
FrontierMath (2025 set)32.4%5.5%
FrontierMath tier 4 (2025 set)12.5%0.0%
Coding
SWE-bench Verified73.6%48.5%
Aider Polyglot88.0%52.4%
LiveCodeBench85.9%54.7%
WeirdML60.7%39.0%
Agents
DTBench90.7%68.3%
LMCA40.0%25.6%
Professional work
CorpFin v261.1%63.0%
LegalBench86.0%83.1%
MortgageTax65.5%65.9%
TaxEval v273.4%75.1%
Novel problems
ARC-AGI-29.9%0.4%
ARC-AGI-165.7%5.5%
Task length · METR42 sec

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

This page as a chartDownload PNG