AIEvals

Models

Claude Fable 5

Specifications and API prices for Claude Fable 5, recorded by Artificial Analysis.

AnthropicReleased 9 June 2026Context 1M tokensOutput Not published in the checked sourcePrivate weights$10.00 in / $50.00 out per Mtok

Family Claude Fable58 benchmarks with a published result4 independent first places among the 46 models this site listsRead 17 October 2026

Where it stands

Rank compares tracked models with a score. Published shows the source’s rank and field size, when available.

BenchmarkScoreRankPublishedLeaderBehind
Vals IndexVals AI61.39%7of 337of 41Gemini 4 Argon7.51 pts
Artificial Analysis Intelligence IndexArtificial Analysis508of 42no list readClaude Opus 5.58 pts
Terminal-Bench 4.0Artificial Analysis42.42%9of 40no list readClaude Sonnet 5.521.22 pts
SWE-bench (Verified)Vals AI95.00%7of 237of 88Claude Opus 52.00 pts
LiveCodeBenchVals AI89.78%2of 212of 143Claude Fable 5.10.74 pts
GPQA DiamondVals AI93.18%9of 2310of 138Gemini 3.1 Pro2.27 pts
MMLU ProVals AI91.50%3of 243of 138Claude Fable 5.10.88 pts
MMMU ProVals AI89.31%3of 203of 93Claude Fable 5.11.33 pts
Vibe Code Bench v1.1Vals AI90.35%3of 383of 106Claude Sonnet 5.52.04 pts
ProgramBenchVals AI2.00%6of 296of 54Claude Opus 5.516.50 pts
τ³-BankingArtificial Analysis38.14%20of 30no list readQwen3.8 Max13.20 pts
IFBenchArtificial Analysis63.47%8of 10no list readGrok 4 20 030919.46 pts
AA-AnalystAgentArtificial Analysis48.75%6of 18no list readGemini 3.7 Flash11.25 pts
MLCR-AAArtificial Analysis64.44%4of 37no list readClaude Sonnet 5.510.56 pts
Vals RSI IndexVals AI24.32%8of 208of 23Claude Opus 5.512.99 pts
τ²-Bench TelecomArtificial Analysis98.54%1of 10no list readJt 35b Flash0.58 pts
Harvey LAB-AAArtificial Analysis93.56%4of 20no list readMuse Spark 1.3 Max1.92 pts
EnterpriseOps-Gym-AAArtificial Analysis51.12%1of 22no list readClaude Fable 5—
AutomationBench-AAArtificial Analysis54.07%27of 40no list readGemini 4 Argon23.44 pts
SciCodeArtificial Analysis61.00%4of 40no list readClaude Opus 5.55.90 pts
GDP.pdfArtificial Analysis24.00%9of 40no list readGPT-6.1 Sol7.00 pts
CritPtArtificial Analysis28.57%10of 42no list readGPT-5.6 Sol3.72 pts
AA-OmniscienceArtificial Analysis43.34of 42no list readClaude Opus 5.53.12
AA-LCR v1.1Artificial Analysis82.33%18of 42no list readKimi K36.34 pts
Terminal-Bench 2.1Vals AI80.52%10of 3510of 75Claude Opus 5.57.12 pts
ProofBench v1.1Vals AI95.00%8of 349of 46Claude Sonnet 5.55.00 pts
LegalBenchVals AI88.56%1of 271of 149Claude Fable 5—
Harvey's Legal Agent BenchmarkVals AI11.25%7of 379of 73Muse Spark 1.214.17 pts
Vals Multimodal IndexVals AI74.15%1of 171of 33Claude Fable 5—
Code MigrationVals AI55.06%8of 378of 71Claude Sonnet 5.514.77 pts
EMBVals AI73.67%5of 375of 69Claude Fable 5.13.00 pts
Finance Agent (v2)Vals AI56.31%12of 3715of 73Gemini 4 Argon9.09 pts
Legal Research BenchVals AI49.52%6of 376of 72Muse Spark 1.3 Max5.77 pts
MedCodeVals AI56.07%4of 354of 104Claude Opus 57.50 pts
MedScribeVals AI88.52%9of 3610of 106Claude Opus 5.52.91 pts
Public Benefits Bench v1.1Vals AI70.43%4of 294of 45Claude Opus 56.50 pts
SAGEVals AI51.89%6of 329of 90Claude Opus 4.74.21 pts
Tax Agent BenchVals AI69.82%9of 3710of 64Claude Fable 5.17.82 pts
DeepSWE v1.1Datacurve69.91%5of 205of 28GPT-6 Astra4.21 pts
GDPval-AA v2.1Artificial Analysis1,595.4119of 42no list readClaude Opus 5.5250.76
AA-Briefcase v1.1Artificial Analysis1,540.7913of 41no list readClaude Opus 5.5281.06
Humanity's Last ExamArtificial Analysis55.47%4of 42no list readClaude Opus 5.55.88 pts

Nearby on Vals Index

The two models above and below on the Vals Index (Vals AI), on every column where this model has a published number.

ModelVals Index%Artificial Analysis Intelligence IndexindexTerminal-Bench 4.0%SWE-bench%LiveCodeBench%GPQA Diamond%MMLU Pro%MMMU Pro%Vibe Code Bench v1.1%ProgramBench%
Claude Opus 5Anthropic63.67%5148.99%97.00%89.03%93.43%91.59%89.88%88.40%3.00%
GPT-6 AstraOpenAI63.13%5359.09%–––––89.59%5.50%
Claude Fable 5Anthropic61.39%5042.42%95.00%89.78%93.18%91.50%89.31%90.35%2.00%
GPT-6.1 SolOpenAI61.15%5256.06%–––––88.93%–
Muse Spark 1.3 MaxMeta58.16%4833.33%–––––85.86%–
Bold marks a figure better than this model’s on the same column, whichever model it belongs to. Ranks cover the models shown from the stated source.

Cost and speed, as published

Additional measurements from Artificial Analysis’ own row. Read 30 September 2026.

As publishedCost per taskOutput speed
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)$8.7562.9 tok/s

Results by benchmark

Independent and vendor results retain their own harnesses and effort settings.

BenchmarkIndependent runAs reported by a publisher
τ³-BankingTool use38.14%Artificial Analysisnot published
IFBenchInstruction following63.47%Artificial Analysisnot published
AA-AnalystAgentData analysis48.75%Artificial Analysisnot published
MLCR-AAMedical reasoning64.44%Artificial Analysisnot published
Vals RSI IndexAutonomous research24.32%Vals AInot published
τ²-Bench TelecomTool use98.54%Artificial Analysisnot published
Harvey LAB-AALegal reasoning93.56%Artificial Analysisnot published
EnterpriseOps-Gym-AAEnterprise workflows51.12%Artificial Analysisnot published
PLawBenchProfessional knowledgeno published result
77.7%PLawBenchZ.ai · 12 October 2026Rank 1 of 5 in this publisher table
AutomationBench-AAWorkflow automation54.07%Artificial Analysisnot published
SciCodeScientific coding61.00%Artificial Analysisnot published
GDP.pdfDocument reasoning24.00%Artificial Analysisnot published
CritPtScientific reasoning28.57%Artificial Analysisnot published
AA-OmniscienceFactual knowledge43.3Artificial Analysisnot published
AA-LCR v1.1Long-context reasoning82.33%Artificial Analysisnot published
Terminal-Bench 4.0Agentic coding42.42%Artificial Analysis
41.41%Vals AI
44.5%Terminal-Bench 4.0OpenAI · 3 September 2026Rank 4 of 6 in this publisher table
Terminal-Bench 2.1Agentic coding80.52%Vals AI
84.64%Artificial Analysis
88.0%Terminal Bench 2.1Z.ai · 14 August 2026Rank 4 of 8 in this publisher table
SWE-bench (Verified)Software engineering95.00%Vals AInot published
LiveCodeBenchCode generation89.78%Vals AInot published
GPQA DiamondScience reasoning93.18%Vals AI
92.63%Artificial Analysis
92.6%GPQA DiamondOpenAI · 3 September 2026Rank 6 of 6 in this publisher table
MMLU ProKnowledge91.50%Vals AInot published
MMMU ProMultimodal reasoning89.31%Vals AInot published
Vibe Code Bench v1.1Product coding90.35%Vals AInot published
Terminal-Bench Science 0.1Agentic science
15.71%Vals AI
21.4%Terminal-Bench Science 0.1OpenAI · 3 September 2026Rank 5 of 5 in this publisher table
ProgramBenchAgentic coding2.00%Vals AI
33.0%ProgramBench (Almost Solved)Z.ai · 14 August 2026Rank 1 of 7 in this publisher table
ProofBench v1.1Mathematics95.00%Vals AInot published
LegalBenchProfessional knowledge88.56%Vals AI
88.6%LegalBenchZ.ai · 12 October 2026Rank 1 of 5 in this publisher table
Harvey's Legal Agent BenchmarkProfessional knowledge11.25%Vals AI
11.3%Harvey's Legal Agent BenchmarkZ.ai · 12 October 2026Rank 3 of 5 in this publisher table
Vals Multimodal IndexComposite74.15%Vals AInot published
Vals IndexComposite61.39%Vals AInot published
Code MigrationAgentic coding55.06%Vals AInot published
EMBFinancial modelling73.67%Vals AInot published
Finance Agent (v2)Financial analysis56.31%Vals AInot published
Legal Research BenchProfessional knowledge49.52%Vals AInot published
MedCodeHealthcare administration56.07%Vals AInot published
MedScribeHealthcare administration88.52%Vals AInot published
Public Benefits Bench v1.1Public services70.43%Vals AInot published
SAGEEducation51.89%Vals AInot published
Tax Agent BenchProfessional knowledge69.82%Vals AInot published
AutomationBenchBusiness workflowsno published result
46.2%AutomationBench (v1.0.6)Z.ai · 14 August 2026Rank 3 of 8 in this publisher table
17.4%AutomationBench 1.0.6OpenAI · 3 September 2026Rank 5 of 5 in this publisher table
DeepSWE v1.1Software engineering69.91%Datacurve
69.7%DeepSWE (v1.1)Z.ai · 14 August 2026Rank 2 of 8 in this publisher table
69.9%DeepSWE v1.1OpenAI · 3 September 2026Rank 5 of 6 in this publisher table
FrontierCode v1.1Agentic codingno published result
53.5FrontierCode 1.1 MainOpenAI · 3 September 2026Rank 1 of 6 in this publisher table
GDPval-AA v2.1Knowledge work1,595.41Artificial Analysisnot published
AA-Briefcase v1.1Knowledge work1,540.79Artificial Analysisnot published
Humanity's Last ExamMultidisciplinary reasoning55.47%Artificial Analysis
63.9%HLE w/ ToolsZ.ai · 14 August 2026Rank 2 of 8 in this publisher table
63.8%Humanity's Last Exam (with tools)OpenAI · 3 September 2026Rank 2 of 4 in this publisher table
ARC-AGI-2Abstract reasoningno published result
89.2%ARC-AGI-2OpenAI · 3 September 2026Rank 5 of 5 in this publisher table
FrontierMath Tier 4 (v2)Mathematicsno published result
90.2%FrontierMath Tier 4 (v2)OpenAI · 3 September 2026Rank 2 of 5 in this publisher table
Agents' Last ExamComputer useno published result
23.8%Agents' Last Exam (ALE-CLI)Z.ai · 14 August 2026Rank 7 of 8 in this publisher table
48.7%Agents' Last ExamOpenAI · 3 September 2026Rank 4 of 4 in this publisher table
Artificial Analysis Intelligence IndexComposite50Artificial Analysisnot published
Terminal-Bench 3.0Agentic codingno published result
33.7%Terminal Bench 3.0Z.ai · 14 August 2026Rank 2 of 6 in this publisher table
CyberGymSecurityno published result
83.8%CyberGymZ.ai · 14 August 2026Rank 2 of 8 in this publisher table
ExploitGymSecurityno published result
181ExploitGym (2h)Z.ai · 14 August 2026Rank 2 of 7 in this publisher table
247ExploitGym (6h)Z.ai · 14 August 2026Rank 2 of 7 in this publisher table
FrontierSWEAgentic researchno published result
47.0%FrontierSWE v2Anthropic · 7 October 2026Rank 1 of 3 in this publisher table
88.2%FrontierSWEZ.ai · 14 August 2026Rank 1 of 4 in this publisher table
SWE-MarathonAgentic codingno published result
33.1%SWE-Marathon (v1.1)Z.ai · 14 August 2026Rank 5 of 6 in this publisher table
PostTrainBenchAI researchno published result
41.8%PostTrainBenchZ.ai · 14 August 2026Rank 1 of 6 in this publisher table
ExploitBenchSecurityno published result
78.0%ExploitBenchZ.ai · 14 August 2026Rank 1 of 7 in this publisher table
Toolathlon VerifiedAgentic toolingno published result
74.7%Toolathlon VerifiedZ.ai · 14 August 2026Rank 4 of 8 in this publisher table
GDPval-AA v2Knowledge workno published result
1,743GDPval-AA v2Z.ai · 14 August 2026Rank 2 of 8 in this publisher table
Independent column: the suite named under each figure, one harness per suite. Publisher column: each lab’s own table, reproduced on the benchmark’s page. Publisher ranks compare reported scores in the same table row only; ties share a rank and unreported cells are excluded.

Sources

Read on 17 October 2026.