The Model Index
Up to four models

Claude Opus 5 vs Claude Opus 4.8

On the index (GPT-4 = 100), Claude Opus 5 scores 185 and Claude Opus 4.8 170. 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 Claude Opus 5
Presets
Claude Opus 5
185169–201
WeightsClosed
ReleasedJul 2026
Price in / out$5.0 / $25
Context1M
Core tests taken6 of 9
Claude Opus 4.8
170154–186
WeightsClosed
ReleasedMay 2026
Price in / out$5.0 / $25
Context1M
Core tests taken6 of 9

Claude Opus 5 leads Claude Opus 4.8 by 15 points, ahead on 6 of 6 shared core tests, at a similar price.

Test by testthird-party results

Claude Opus 5Claude Opus 4.8
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)LiveBench Data AnalysisLiveBench Data AnalysisLiveBench IFLiveBench IFLiveBench LanguageLiveBench LanguageLiveBench ReasoningLiveBench ReasoningGPQA Diamond (Vals AI)GPQA Diamond (Vals AI)MMLU ProMMLU ProMMMU ProMMMU ProEBR-benchEBR-benchSimpleBenchSimpleBenchSimpleQA VerifiedSimpleQA VerifiedMATHSFrontierMath (tiers 1-3)FrontierMath (tiers 1-3)AIME-style maths (OTIS mock)AIME-style maths (OTIS mock)LiveBench MathematicsLiveBench MathematicsChess puzzlesChess puzzlesFrontierMath (tier 4)FrontierMath (tier 4)CODINGTerminal-Bench 4.0Terminal-Bench 4.0LiveBench Agentic CodingLiveBench Agentic CodingLiveBench CodingLiveBench CodingCode MigrationCode MigrationLiveCodeBenchLiveCodeBenchProgramBenchProgramBenchSWE-benchSWE-benchVibe Code Bench v1.1Vibe Code Bench v1.1DeepSWEDeepSWEMirrorCodeMirrorCodeWeirdMLWeirdMLAGENTSAPEX-AgentsAPEX-AgentsSkillsBenchSkillsBenchTerminal-Bench-ScienceTerminal-Bench-ScienceDTBenchDTBenchFurniture assemblyFurniture assemblyLMCALMCAMystery game puzzlesMystery game puzzlesPROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)CorpFin v2CorpFin v2EMBEMBLegalBenchLegalBenchLegal Research BenchLegal Research BenchMedCodeMedCodeMedScribeMedScribeMortgageTaxMortgageTaxPublic Benefits Bench v1.1Public Benefits Bench v1.1Tax Agent BenchTax Agent BenchTaxEval v2TaxEval v2NOVEL PROBLEMSARC-AGI-2ARC-AGI-2ARC-AGI-1ARC-AGI-1ARC-AGI-3ARC-AGI-3
Every result as a table
TestClaude Opus 5Claude Opus 4.8
Reasoning & knowledge
GPQA Diamond (Epoch AI)93.9%91.0%
LiveBench Data Analysis74.6%66.0%
LiveBench IF63.8%72.0%
LiveBench Language88.7%79.7%
LiveBench Reasoning91.2%89.2%
GPQA Diamond (Vals AI)93.4%92.4%
MMLU Pro91.6%89.6%
MMMU Pro89.9%86.6%
EBR-bench45.7%28.6%
SimpleBench80.6%64.8%
SimpleQA Verified59.9%53.0%
Maths
FrontierMath (tiers 1-3)85.6%80.0%
AIME-style maths (OTIS mock)98.9%98.3%
LiveBench Mathematics95.7%94.3%
Chess puzzles43.0%52.0%
FrontierMath (tier 4)73.2%56.1%
Coding
Terminal-Bench 4.053.5%23.2%
LiveBench Agentic Coding65.2%50.5%
LiveBench Coding81.5%81.8%
Code Migration57.5%47.3%
LiveCodeBench89.0%87.8%
ProgramBench3.0%1.0%
SWE-bench97.0%88.6%
Vibe Code Bench v1.188.4%82.7%
DeepSWE73.7%59.0%
MirrorCode87.8%65.6%
WeirdML91.8%82.9%
Agents
APEX-Agents65.8%48.9%
SkillsBench60.4%59.2%
Terminal-Bench-Science27.1%4.3%
DTBench97.9%94.9%
Furniture assembly60.8%42.5%
LMCA64.5%57.5%
Mystery game puzzles59.0%36.0%
Professional work
Finance Agent (v2)58.6%53.9%
CorpFin v273.2%66.7%
EMB73.6%69.4%
LegalBench87.0%83.6%
Legal Research Bench55.3%43.8%
MedCode63.6%53.2%
MedScribe91.0%85.8%
MortgageTax72.1%69.9%
Public Benefits Bench v1.176.9%68.1%
Tax Agent Bench47.1%30.0%
TaxEval v275.1%75.6%
Novel problems
ARC-AGI-290.4%72.1%
ARC-AGI-197.5%92.5%
ARC-AGI-330.2%1.5%

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

This page as a chartDownload PNG