The Model Index
Up to four models

Claude Opus 5 vs Inkling

On the index (GPT-4 = 100), Claude Opus 5 scores 185 and Inkling 149. 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
Inkling
149133–165
WeightsOpen
ReleasedJul 2026
Price in / out$1.0 / $4.05
Context524K
Core tests taken6 of 9

Claude Opus 5 leads Inkling by 36 points, ahead on 6 of 6 shared core tests, at 5.7× the price.

Test by testthird-party results

Claude Opus 5Inkling
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)TMLiveBench Data AnalysisLiveBench Data AnalysisTMLiveBench IFLiveBench IFTMLiveBench LanguageLiveBench LanguageTMLiveBench ReasoningLiveBench ReasoningTMGPQA Diamond (Vals AI)GPQA Diamond (Vals AI)TMMMLU ProMMLU ProTMMMMU ProMMMU ProTMSimpleBenchSimpleBenchTMSimpleQA VerifiedSimpleQA VerifiedTMMATHSFrontierMath (tiers 1-3)FrontierMath (tiers 1-3)TMAIME-style maths (OTIS mock)AIME-style maths (OTIS mock)TMLiveBench MathematicsLiveBench MathematicsTMProofBench v1.1ProofBench v1.1TMChess puzzlesChess puzzlesTMFrontierMath (tier 4)FrontierMath (tier 4)TMCODINGTerminal-Bench 4.0Terminal-Bench 4.0TMLiveBench Agentic CodingLiveBench Agentic CodingTMLiveBench CodingLiveBench CodingTMCode MigrationCode MigrationTMCyberBench v1.1CyberBench v1.1TMIOIIOITMLiveCodeBenchLiveCodeBenchTMProgramBenchProgramBenchTMSWE-benchSWE-benchTMVibe Code Bench 1-100Vibe Code Bench 1-100TMVibe Code Bench v1.1Vibe Code Bench v1.1TMFrontierSWEFrontierSWETMWeirdMLWeirdMLTMAGENTSAPEX-AgentsAPEX-AgentsTMSkillsBenchSkillsBenchTMSRE BenchSRE BenchTMDTBenchDTBenchTMLMCALMCATMPROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)TMCorpFin v2CorpFin v2TMEMBEMBTMLegalBenchLegalBenchTMLegal Research BenchLegal Research BenchTMMedCodeMedCodeTMMedScribeMedScribeTMMortgageTaxMortgageTaxTMPublic Benefits Bench v1.1Public Benefits Bench v1.1TMTax Agent BenchTax Agent BenchTMTaxEval v2TaxEval v2TMNOVEL PROBLEMSARC-AGI-2ARC-AGI-2TMARC-AGI-1ARC-AGI-1TM
Every result as a table
TestClaude Opus 5Inkling
Reasoning & knowledge
GPQA Diamond (Epoch AI)93.9%88.3%
LiveBench Data Analysis74.6%72.8%
LiveBench IF63.8%70.1%
LiveBench Language88.7%73.5%
LiveBench Reasoning91.2%78.3%
GPQA Diamond (Vals AI)93.4%87.1%
MMLU Pro91.6%86.3%
MMMU Pro89.9%78.5%
SimpleBench80.6%50.0%
SimpleQA Verified59.9%40.3%
Maths
FrontierMath (tiers 1-3)85.6%33.3%
AIME-style maths (OTIS mock)98.9%88.9%
LiveBench Mathematics95.7%88.4%
ProofBench v1.199.0%0.0%
Chess puzzles43.0%21.0%
FrontierMath (tier 4)73.2%4.9%
Coding
Terminal-Bench 4.053.5%0.5%
LiveBench Agentic Coding65.2%49.4%
LiveBench Coding81.5%71.0%
Code Migration57.5%11.8%
CyberBench v1.165.4%66.8%
IOI84.3%14.9%
LiveCodeBench89.0%85.5%
ProgramBench3.0%0.0%
SWE-bench97.0%77.6%
Vibe Code Bench 1-10028.5%7.3%
Vibe Code Bench v1.188.4%19.2%
FrontierSWE52.0%4.1%
WeirdML91.8%32.3%
Agents
APEX-Agents65.8%33.8%
SkillsBench60.4%26.1%
SRE Bench12.2%0.0%
DTBench97.9%87.5%
LMCA64.5%37.6%
Professional work
Finance Agent (v2)58.6%46.6%
CorpFin v273.2%68.6%
EMB73.6%38.4%
LegalBench87.0%83.0%
Legal Research Bench55.3%28.4%
MedCode63.6%41.2%
MedScribe91.0%85.4%
MortgageTax72.1%64.0%
Public Benefits Bench v1.176.9%58.5%
Tax Agent Bench47.1%15.3%
TaxEval v275.1%75.3%
Novel problems
ARC-AGI-290.4%36.5%
ARC-AGI-197.5%79.5%

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

This page as a chartDownload PNG