Atlas

Compare Models

Claude Sonnet 5.5 vs Inkling 2 Small

Claude Sonnet 5.5 leads by 6.5 capability points (≈95% interval +4.1 to +8.8): a clear lead over Inkling 2 Small · P(higher) ≈ 100%

Wins 58 of 87 shared benchmarks (20 lost, 9 tied)

Capability vs. relative cost
Claude Sonnet 5.5Inkling 2 SmallClosed 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 Sonnet 5.5vsInkling 2 Smallcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins58 – 20 · 87 shared, 9 tied

Comparing matching setups only (50 benchmarks excluded) · suites counted once
← Claude Sonnet 5.5Inkling 2 Small →

Claude Sonnet 5.5vsInkling 2 Smallwins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Surge AI and 11 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Inkling 2 Small minus Claude Sonnet 5.5.

Terminal-Bench 4.0 (Artificial Analysis)Code63.613.1−50.5
HLAB-Vals-1.1 - Energy Natural ResourcesLegal0.050.0+50.0
Terminal-Bench-Science 0.1Science53.35.7−47.6
Terminal-Bench Science 0.1 (Vals)Science47.12.9−44.3
HLAB-Vals-1.1 - Litigation Dispute ResolutionLegal0.040.0+40.0
sudo L7 (October 2026 rescoring)Code50.614.1−36.5
AutomationBench-AA FinanceAgents79.443.9−35.4
HLAB-Vals-1.1 - Antitrust CompetitionLegal10.040.0+30.0
HLAB-Vals-1.1 - Arbitration International Dispute ResolutionLegal0.030.0+30.0
HANDBOOK.md Agents v1.1Professional Work47.717.8−29.9
DAYJOB: Finance (October 2026 rescoring)Finance & Accounting40.514.3−26.2
AA-3D v1.0 EloCode12041027−25.6
ChartographyMultimodal61.636.7−24.9
DeepFinanceFinance & Accounting54.629.8−24.8
Humanity's Last Exam DiamondGeneral QA53.328.7−24.6
AutomationBench-AA Operations, HR, Marketing, Sales and SupportAgents70.146.0−24.1
MysteryMechanismScience49.125.2−23.9
APEX-Agents 1.1Professional Work83.760.9−22.8
Arena Code WebDev — ReactCode18221659−21.8
AutomationBench-AA Support and OperationsAgents76.554.8−21.7
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine41.720.6−21.1
FrontierCode 2.0 - Pass RateCode46.425.5−20.8
CritPt v2 (agentic, avg@5)Science75.755.7−20.0
HLAB-Vals-1.1 - Corporate GovernanceLegal10.030.0+20.0
HLAB-Vals-1.1 - Funds Asset ManagementLegal10.030.0+20.0
HLAB-Vals-1.1 - International Trade SanctionsLegal0.020.0+20.0
HLAB-Vals-1.1 - TaxLegal0.020.0+20.0
FrontierCode 2.0 - ScoreCode42.222.6−19.7
Artificial Analysis Engineering IndexIndexes58.538.9−19.6
ComplexConstraints (November 2026 rescoring)Chat & Writing67.247.6−19.6
VIALS-AA (pass^3)Multimodal59.642.9−16.8
AA-Briefcase v1.1 EloAgents18161527−15.9
SWE-bench Multimodal (Vals v1, 480 Tasks)Code58.343.1−15.2
Artificial Analysis Strategy & Ops IndexIndexes58.143.5−14.6
SWE-bench VerifiedCode96.782.2−14.5
Artificial Analysis Healthcare & Medical IndexIndexes54.240.8−13.5
Arena Code WebDev — HTMLCode17521660−13.0
Excel Modeling Benchmark v1.1 LBOFinance & Accounting68.755.9−12.8
SWE-Marathon v1.1Code51.338.8−12.4
Artificial Analysis Finance & Accounting IndexIndexes55.844.3−11.4
Vibe Code Bench v1.1Code92.481.5−10.8
Excel Modeling Benchmark v1.1 DCFFinance & Accounting73.062.2−10.7
Artificial Analysis Economics IndexIndexes54.143.4−10.7
Artificial Analysis Legal IndexIndexes52.241.6−10.6
HLAB-Vals-1.1 - Banking FinanceLegal0.010.0+10.0
HLAB-Vals-1.1 - Bankruptcy RestructuringLegal10.020.0+10.0
HLAB-Vals-1.1 - Emerging Companies Venture CapitalLegal10.020.0+10.0
HLAB-Vals-1.1 - Healthcare Life SciencesLegal20.030.0+10.0
HLAB-Vals-1.1 - ImmigrationLegal20.030.0+10.0
HLAB-Vals-1.1 - Intellectual PropertyLegal30.020.0−10.0
HLAB-Vals-1.1 - Real EstateLegal20.030.0+10.0
HLAB-Vals-1.1 - Structured Finance SecuritizationLegal30.020.0−10.0
SWE-bench MultilingualCode90.981.2−9.7
Excel Modeling Benchmark v1.1 TemplateFinance & Accounting80.971.6−9.3
GDPval-AA v2.1Professional Work18381630−9.1
SciCode 1.0.1 (Artificial Analysis)Code61.052.3−8.7
Excel Modeling Benchmark v1.1 M&AFinance & Accounting82.373.7−8.6
HLAB-Vals-1.1 - Criteria Pass RateLegal85.993.7+7.8
VIALS-AA (average of 3)Multimodal65.457.8−7.7
AA-Omniscience v1.1 IndexIndexes30.516.1−7.2
Excel Modeling Benchmark v1.1 CompsFinance & Accounting79.473.1−6.3
AA-AnalystAgent (pass^5)Professional Work57.551.3−6.2
ProgramBenchCode6.50.5−6.0
Terminal-Bench 2.1Code87.381.3−6.0
SWE Atlas v1.1 - OverallCode67.561.9−5.6
Excel Modeling Benchmark v1.1 DataroomFinance & Accounting87.783.4−4.3
Excel Modeling Benchmark v1.1 ScratchFinance & Accounting70.867.1−3.7
SAGE v1.1 - Partial Differential EquationsMath91.194.0+2.9
APEX-Accounting 1.1Finance & Accounting64.562.0−2.5
ARC-AGI-2General QA83.981.7−2.2
DeepSWE v1.1Code70.572.1+1.6
Excel Modeling Benchmark v1.1 OperatingFinance & Accounting71.470.0−1.5
VIALS-AA (pass@3)Multimodal70.872.0+1.2
EnterpriseBench: CoreCraftProfessional Work70.871.8+1.0
GDP.pdf (Artificial Analysis, All-pass)Professional Work25.825.2−0.6
Excel Modeling Benchmark v1.1 MiscFinance & Accounting68.868.3−0.6
SAGE v1.1 - Linear AlgebraMath80.380.5+0.2
SAGE v1.1 - Computer ScienceMath91.491.2−0.2
ARC-AGI-1 Semi-PrivateGeneral QA96.596.50.0
HLAB-Vals-1.1 - Capital MarketsLegal20.020.00.0
HLAB-Vals-1.1 - Corporate M&ALegal20.020.00.0
HLAB-Vals-1.1 - Data Privacy CybersecurityLegal20.020.00.0
HLAB-Vals-1.1 - Employment LaborLegal0.00.00.0
HLAB-Vals-1.1 - Environmental ESGLegal0.00.00.0
HLAB-Vals-1.1 - InsuranceLegal0.00.00.0
HLAB-Vals-1.1 - Trusts Estates Private ClientLegal20.020.00.0
HLAB-Vals-1.1 - White Collar Defense InvestigationsLegal0.00.00.0

Claude Sonnet 5.5vsInkling 2 Smallnormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Surge AI and 11 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…