Atlas

Compare Models

Claude Fable 5.1 vs GPT-6.1 Sol

GPT-6.1 Sol leads by 0.6 capability points (≈95% interval −2.0 to +3.1): statistically tied · P(higher) ≈ 66%

Claude Fable 5.1 wins 114 of 211 shared benchmarks (91 lost, 6 tied)

Capability vs. relative cost
Claude Fable 5.1GPT-6.1 SolClosed weightsOpen weights

Each line runs through one model’s priced reasoning settings, cheapest to most expensive; grey points are other models’ priced settings in the same range. Results whose reasoning effort the source didn’t report are left out wherever a model has named effort levels. Capability and relative cost are the published estimates (relative cost is cost per task as a multiple of GPT-5 (high)); hover a setting, or use the arrow keys, to set it beside the other model’s same setting.

Claude Fable 5.1vsGPT-6.1 Solcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins114 – 91 · 211 shared, 6 tied

Comparing matching setups only (92 benchmarks excluded) · suites counted once
← Claude Fable 5.1GPT-6.1 Sol →

Claude Fable 5.1vsGPT-6.1 Solwins by categoryData Nov 15, 2026Data: Mercor, Artificial Analysis, Vals AI and 55 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-6.1 Sol minus Claude Fable 5.1.

VCT-Refusal (CAIS)Safety98.916.1−82.8
AA-PersonalityBench Informativeness indexChat & Writing96.227.7−69.2
Concept Synth INDUCTION Challenge64 Round 1General QA34.495.3+60.9
HPCT Refusal (CAIS)Safety100.042.0−58.0
Concept Synth INDUCTION Challenge100 Round 1General QA33.090.0+57.0
ActiveVisionMultimodal15.371.8+56.5
Olam Social Poker EloGames14941543+53.3
KernelBench Mega Normalized (RTX PRO 6000)Code64.714.1−50.6
Clipping Bench v1.0Games10685707+49.9
AA-PersonalityBench Expressiveness indexChat & Writing98.150.5−48.1
Design Creativity BenchCode48.663.7+47.9
AA-PersonalityBench Conviction indexChat & Writing10054.3−46.2
Medical Long Context Reasoning (MLCR-AA)Health & Medicine71.133.9−37.2
BioSecBench-Refusal V1 Clean RerunsSafety5.637.6+32.0
AA-PersonalityBench Warmth indexChat & Writing94.365.7−28.8
Clocktower BenchGames15671764+28.3
ChessBench EloGames10951562+27.8
EnterpriseBench: CoreCraftProfessional Work77.450.3−27.1
Terminal-Bench 2.1 Extended (Mercor held-out)Code19.545.8+26.3
ChessBench CoherenceGames61.987.5+25.6
AA-PersonalityBench Imagination indexChat & Writing86.761.9−25.0
ChessBenchGames45.368.7+23.3
Mystery Game PuzzlesGames58.080.0+22.0
SpireClimbGames2.023.8+21.8
AA-PersonalityBench Organization indexChat & Writing60.081.0+21.2
SM BenchSafety89.069.0−20.1
Chess Puzzles (task version 2.0)Games48.068.0+20.0
LisanBench Path LengthGames151008736−19.8
HANDBOOK.md Agents v1.1Professional Work47.128.6−18.5
BrowseComp (Mercor 130-question subset)Search78.395.4+17.1
APEX-SWECode63.646.9−16.7
SWE-bench Verified Extended (Mercor held-out)Code61.345.3−16.0
Tax Agent BenchFinance & Accounting77.962.3−15.6
Agent Red Teaming k=100 (CAIS)Safety34.218.8+15.5
EnigmaEvalGeneral QA44.259.5+15.3
SnorkelLegalLegal35.920.7−15.2
ClockBench AccuracyMultimodal59.474.4+15.0
WellnessBenchSafety51.066.0+15.0
BioSecBench-Structure V1 (58 evaluations)Science6.321.3+14.9
TextQuests Game ProgressGames56.671.4+14.8
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting77.063.2−13.8
MMMU-Pro Extended (Mercor held-out)Multimodal47.661.3+13.7
BrokenArXiv 06/2026Math86.6100.0+13.4
Visa Agentic Commerce BenchmarkAgents75.662.2−13.3
SnorkelFinance 2.0Finance & Accounting19.66.6−13.0
MASKSafety85.998.3+12.4
SpatialViz-BenchMultimodal77.990.3+12.3
FrontierMath Tier 4 (v2) - PrivateMath87.8100.0+12.2
HardGeoBench Avg DistanceMultimodal8451239−12.1
OpenWorldEngineer (Mean@8)Code29.141.2+12.0
BrokenArXiv 05/2026Math78.090.0+12.0
BullshitBench v2 LeaderboardSafety77.065.0−12.0
Finance Agent v2 - Disclosure AnalysisFinance & Accounting70.358.4−11.9
NoCoT-Bench NCRI 15.2General QA128149+11.5
Agent Initiative BenchmarkAgents7.58.7+11.5
HieroglyphBenchMultimodal20.431.4+11.0
SWE-TogetherCode69.358.3−11.0
τ³-Banking (Mercor run, bm25_grep retrieval)Agents54.043.0−11.0
Machiavelli Deception (CAIS)Safety82.478.8+10.9
SpeciEval (Refusal-Aware Retries, Historical)Safety87.297.4+10.1
SpeciEval (Pooled English Runs)Safety87.297.4+10.1
ChessBench AccuracyGames71.380.7+10.1
FrontierCUA pass@1Agents36.746.7+10.0
Furniture Assembly BenchmarkMultimodal70.080.0+10.0
HardGeoBench <=100miMultimodal53.843.8−10.0
HardGeoBench <=10miMultimodal45.035.0−10.0
GBENCH One-shot Coding GScoreGames69.379.0+9.7
ComplexConstraints (November 2026 rescoring)Chat & Writing60.470.1+9.7
Finance Agent v2 - AdjustmentsFinance & Accounting54.444.7−9.7
AA-PersonalityBench Slop indexChat & Writing23.814.3−9.6
SciCode (main problems)Science24.214.6−9.6
SWE-bench VerifiedCode96.687.3−9.3
BuseyBench SVGMultimodal7.08.0+9.2
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science3.612.4+8.8
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal17.48.8−8.6
KernelBench CUDA Normalized (RTX PRO 6000)Code51.643.2−8.4
SnorkelManufacturingProfessional Work11.319.6+8.3
SWE Atlas v1.1 - OverallCode74.165.9−8.2
RiemannBench (November 2026 rescoring)Math71.279.2+8.0
DeepSWE v1.1Code67.375.2+7.9
ERQAMultimodal70.378.2+7.9
Finance Agent v2 - All PassFinance & Accounting47.839.9−7.9
Arena Alignment IndexSafety80.187.9+7.8
Immigration BenchLegal75.467.8−7.6
Gray Swan IPI Attack Success (k=15)Safety1.08.6−7.6
TextQuests Average Cumulative HarmSafety27.129.3−7.6
ArXivMath 05/2026Math87.595.0+7.5
CVDP (517-problem set)Code62.655.1−7.5
AA-AnalystAgent (pass^5)Professional Work57.550.0−7.5
DAYJOB: Data ScienceProfessional Work26.018.7−7.3
SciCode 1.0.1 (Artificial Analysis)Code63.155.8−7.3
Financial Coaching BenchFinance & Accounting75.768.5−7.2
GeneBench-Pro Public Release SubsetScience26.033.0+7.0
Roboflow Vision Evals (October 2026, five tasks)Multimodal74.781.7+7.0
HealthBenchHealth & Medicine63.956.9−7.0
GDPval-AA v2.1Professional Work17651607−6.9
SWE Atlas Codebase QnA Extended (Mercor held-out)Code38.031.3−6.7
FrontierCode 2.0 (Artificial Analysis)Code30.336.7+6.4
Finance Agent v2 - PrecedentsFinance & Accounting32.225.8−6.4
AA-Briefcase v1.1 EloAgents16711555−6.3
HardGeoBench Country AccuracyMultimodal58.852.5−6.2
SWE-bench MultilingualCode91.184.9−6.2
LisanBench Difficulty-WeightedGames45734004−6.1
Excel Modeling BenchmarkFinance & Accounting76.770.8−5.9
AA-PersonalityBench Transparency indexChat & Writing98.192.4−5.8
Excel Modeling Benchmark v1.1Finance & Accounting76.771.1−5.6
DAYJOB: TaxFinance & Accounting9.64.0−5.6
Search ArenaSearch12371230−5.6
Mercor enterprise risk (implicit-risk unsafe rate)Safety77.883.3−5.6
Extended NYT Connections 940 (Quadratic v1)Games90.095.5+5.5
DAYJOB: LawLegal26.531.8+5.3
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting79.774.5−5.2
SWE Atlas - Codebase QnACode43.348.4+5.1
Finance Agent v2 - Market AnalysisFinance & Accounting69.564.5−5.1
ProgramBench (Mercor 100-task subset)Code11.36.3−5.0
Toloka Arena (v3)Agents71.066.0−5.0
ALE-BenchCode28492990+5.0
MedScribeHealth & Medicine91.386.5−4.8
Humanity's Last ExamGeneral QA46.551.3+4.8
Snorkel SWE-bench CLICode14.59.8−4.7
MedCodeHealth & Medicine53.548.8−4.7
IntPhys 2Multimodal67.772.3+4.6
MindCube TinyMultimodal84.589.1+4.6
Professional Reasoning Benchmark - FinanceFinance & Accounting50.846.2−4.6
Terminal-Bench 4.0Code57.953.4−4.5
LMArena Agent Arena Net Improvement (Steering Burden)Agents14.812.5−4.5
FrontierMath ErdősMath0.04.4+4.4
CritPt v2 (agentic, avg@5)Science77.782.0+4.3
Roboflow Vision Evals (September 2026 protocol)Multimodal81.385.5+4.2
ARC-AGI-2General QA90.094.2+4.2
Design Arena Overall Code CategoriesCode13311356+4.1
Finance Agent v2 - Earnings AnalysisFinance & Accounting75.571.5−4.0
TaxCalcBench TY25Finance & Accounting42.038.0−4.0
PCT Political Inconsistency (CAIS)Safety33.437.3−3.9
APEX-Agents 1.1Professional Work77.173.2−3.9
DualEntry Accounting AI BenchmarkFinance & Accounting85.181.2−3.9
Terminal-Bench 2.1Code81.685.4+3.8
Professional Reasoning Benchmark - LegalLegal51.647.9−3.7
SimpleBenchGeneral QA86.682.9−3.7
Code2Video-AA EloCode13251369+3.6
LMArena Text Arena EloChat & Writing15031482−3.6
LMArena Agent Arena Bash RecoveryAgents11.64.5−3.5
FrontierMath Tiers 1-3 (v2) - PrivateMath90.293.7+3.5
AA-LCR v1.1Search87.584.0−3.5
ArXivMath 06/2026Math91.094.4+3.5
Finance Agent v2 - ComparablesFinance & Accounting46.042.6−3.4
Ramp Accounting BenchFinance & Accounting49.746.4−3.3
Westworld Finance Diligence BenchFinance & Accounting58.255.0−3.2
Snorkel Agentic Coding 2.0Code39.642.8+3.2
Conceptual Reasoning IndexIndexes74.972.7−3.1
SimpleQA Verified (Epoch anti-abstention)General QA70.873.9+3.1
SUPERChemScience74.777.8+3.1
CheatBenchSafety45.842.8+3.0
CharXiv Extended (Mercor held-out)Multimodal82.285.1+2.9
EBR-bench v4 (card ban, single agent)Games57.154.3−2.9
ACCoRDGeneral QA83.780.9−2.8
LMCAGeneral QA65.562.8−2.7
DistressBenchSafety71.574.2+2.6
AA-3D v1.0 EloCode11391121−2.6
LiveBench 2.0Agents88.085.4−2.6
LMArena Agent Arena Bash ReliabilityAgents10.15.0−2.6
Snorkel Digital Work IndexIndexes23.621.2−2.4
APEX-Accounting 1.1Finance & Accounting66.764.3−2.4
HoneyBench v0.2 (hack rate)Safety61.859.5+2.4
MissionReportProfessional Work63.361.0−2.3
Finance Agent v2 - Financial ModelingFinance & Accounting25.323.1−2.2
GameBench 2Games94.092.0−2.1
LMArena Agent Arena Praise vs ComplaintAgents31.627.5−2.1
DAYJOB: AccountingFinance & Accounting27.225.2−2.0
DAYJOB: HealthcareHealth & Medicine9.67.6−2.0
MedXpertQA MM (200-question set)Health & Medicine83.785.7+2.0
MMLU ProGeneral QA93.091.0−2.0
SAGEMath48.546.5−2.0
SnorkelRevOpsProfessional Work14.012.1−1.9
DAYJOB: FinanceFinance & Accounting19.818.0−1.8
AdvancedIFChat & Writing84.282.4−1.8
sudo L7 (October 2026 rescoring)Code32.233.9+1.7
Epoch Capabilities Index (ECI)Indexes167169+1.5
LMArena Agent Arena Confirmed SuccessAgents14.411.7−1.4
MysteryMechanismScience47.746.4−1.4
MedXpertQA MM Extended (Mercor held-out)Health & Medicine66.064.7−1.3
HealthBench (length-adjusted, Artificial Analysis)Health & Medicine57.258.5+1.3
GDP.xlsxProfessional Work23.122.0−1.1
DTBenchGeneral QA97.696.5−1.1
DAXBench v2Code91.592.6+1.1
MMMU ProMultimodal86.185.1−1.0
ARC-AGI-1 Semi-PrivateGeneral QA97.598.5+1.0
ProofBench v1.1Math100.099.0−1.0
SnorkelUnderwrite 2.0Finance & Accounting24.025.0+1.0
Ruby LLM Benchmarks — Program FixerCode76.075.0−0.9
CharXiv (descriptive and reasoning combined)Multimodal95.196.0+0.9
HealthBench ProfessionalHealth & Medicine69.268.3−0.9
NoCoT-Bench NCKI (kspine v3)General QA128130+0.9
HiL-Bench (Human-in-Loop Benchmark)Agents61.562.3+0.8
Vision Arena OverallMultimodal12891293+0.8
Snorkel WorkplaceAgentsProfessional Work15.815.0−0.8
LMArena Agent Arena Steering BurdenAgents14.813.4−0.7
ObviousBenchGeneral QA99.3100.0+0.7
CritPt (avg@5)Science31.131.7+0.6
Harvey LAB-AA v1.1 (hallucination-gated)Legal6.46.9+0.6
Prophet Arena Arena Score (Fixed Context)Forecasting85.084.8−0.2
FrontierCode 1.1 Main (Artificial Analysis)Code49.049.1+0.1
LMArena Agent Arena Tool HallucinationAgents0.81.0+0.1
GDPxProfessional Work8.58.4−0.1
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.086.00.0
AIME 2025Math100.0100.00.0
Dual-Arm Sim (task completion)Agents0.00.00.0
Mercor enterprise risk (benign-control completion rate)Safety100.0100.00.0
Mercor enterprise risk (explicit-risk unsafe rate)Safety0.00.00.0
OTIS Mock AIME 2024-2025Math100.0100.00.0
TxBench-Antibody-Development V1 (60 evaluations)Science27.827.80.0

Claude Fable 5.1vsGPT-6.1 Solnormalised score marginsData Nov 15, 2026Data: Mercor, Artificial Analysis, Vals AI and 55 othersVillage LLM Atlasatlas.village.dev

Win counts use each benchmark’s declared score direction. Represented suite slices and the derived Capability score are excluded from headline totals. Missing or incompatible setups are excluded, never counted as losses.
Loading Atlas data…