Atlas

Compare Models

Claude Haiku 5.5 vs Grok 4.6

Claude Haiku 5.5 leads by 1.3 capability points (≈95% interval −0.7 to +3.4): statistically tied · P(higher) ≈ 90%

Grok 4.6 wins 83 of 176 shared benchmarks (76 lost, 17 tied)

Capability vs. relative cost
Claude Haiku 5.5Grok 4.6Closed 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 Haiku 5.5vsGrok 4.6capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins76 – 83 · 176 shared, 17 tied

Comparing matching setups only (23 benchmarks excluded) · suites counted once
← Claude Haiku 5.5Grok 4.6 →

Claude Haiku 5.5vsGrok 4.6wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 35 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Grok 4.6 minus Claude Haiku 5.5.

Medical Long Context Reasoning (MLCR-AA)Health & Medicine63.312.2−51.1
Roboflow Vision Evals — Detection mAP@75 (September)Multimodal54.36.1−48.2
AA-PersonalityBench Imagination indexChat & Writing63.820.0−44.2
Roboflow Vision Evals — Detection mAP@50 (September)Multimodal68.224.0−44.2
Roboflow Vision Evals — Detection mAP@50:95 (September)Multimodal53.09.5−43.5
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal0.040.0+40.0
AA-PersonalityBench Warmth indexChat & Writing69.531.5−38.5
ProofBench v1.1Math89.051.0−38.0
Toloka Arena (v3)Agents21.056.0+35.0
Vending-Bench 2Games36029047+33.0
AutomationBench-AA 1.0.6Agents36.067.0+31.0
AA-PersonalityBench Informativeness indexChat & Writing75.344.8−30.8
LLM Chess EloGames1143443−28.7
BrowseComp (Mercor 130-question subset)Search56.384.0+27.7
AA-PersonalityBench Organization indexChat & Writing41.014.3−26.9
SimpleQA Verified (Epoch anti-abstention)General QA23.849.3+25.5
VoxelBench Text RatingGames16962050+25.2
AA-PersonalityBench Conviction indexChat & Writing67.692.4+25.0
AA-PersonalityBench Slop indexChat & Writing10.535.3+25.0
SWE-bench Verified Extended (Mercor held-out)Code53.378.0+24.7
Eval Connections One-Shot with TrapsGames67.091.0+24.0
Chess Puzzles (task version 2.0)Games55.033.0−22.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Banking FinanceLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal0.020.0+20.0
Harvey HLAB — Criteria Pass RateLegal73.492.5+19.1
FrontierSWE V2Code43.825.3−18.5
SWE-Marathon v1.1Code50.031.9−18.1
NoCoT-Bench NCKI (kspine v3)General QA85.9112+17.9
SAGEMath45.828.9−16.9
HealthBench ProfessionalHealth & Medicine64.848.5−16.3
Terminal-Bench-Science 0.1Science18.12.9−15.2
HardGeoBench Avg DistanceMultimodal13191807−15.0
SM BenchSafety74.389.0+14.7
FrontierMath Tier 4 (v2) - PrivateMath46.331.7−14.6
Extended NYT Connections 940 (Quadratic v1)Games65.780.0+14.3
Vibe Code Bench v1.1Code90.476.2−14.2
MineBench — Bradley-Terry RatingGames15821811+13.9
AA-PersonalityBench Expressiveness indexChat & Writing77.263.8−13.5
AA-AnalystAgent (pass^5)Professional Work53.841.3−12.5
HardGeoBench <=100miMultimodal45.032.5−12.5
Terminal-Bench 4.0 (Artificial Analysis)Code29.317.2−12.1
ChessBench CoherenceGames80.769.2−11.5
Wiz Cyber Model Arena v3.9Cybersecurity57.045.5−11.5
Roboflow Vision Evals — Reasoning, Judged (September)Multimodal74.863.8−11.0
Roboflow Vision Evals (October 2026, five tasks)Multimodal69.659.1−10.5
FrontierCode 2.0 (Artificial Analysis)Code29.919.8−10.2
NoCoT-Bench NCRI 15.2General QA68.386.8+10.1
HardGeoBench <=10miMultimodal30.020.0−10.0
ACCoRDGeneral QA81.571.6−9.9
AA-Omniscience IndexIndexes10.730.5+9.9
Tax Agent BenchFinance & Accounting62.271.4+9.2
FrontierMath Tiers 1-3 (v2) - PrivateMath75.166.0−9.1
APEX-Agents 1.1Professional Work68.377.3+9.0
Roboflow Vision Evals — Reasoning, Strict (September)Multimodal72.863.8−9.0
Humanity's Last Exam DiamondGeneral QA29.621.5−8.1
Terminal-Bench 4.0Code33.325.3−8.0
RedlineBenchLegal40.648.5+7.9
Terminal-Bench 2.1 Extended (Mercor held-out)Code37.745.5+7.8
CyberBench v1.1 OverallCybersecurity75.567.7−7.8
SWE-bench MultilingualCode85.192.7+7.6
Furniture Assembly BenchmarkMultimodal47.540.0−7.5
HardGeoBench Country AccuracyMultimodal51.343.8−7.5
MedXpertQA MM (200-question set)Health & Medicine67.875.2+7.4
Roboflow Vision Evals — Counting, Judged (September)Multimodal73.065.8−7.2
Roboflow Vision Evals — Counting, Strict (September)Multimodal73.065.8−7.2
CursorBench 4.0Code48.441.4−7.0
MedCodeHealth & Medicine51.544.7−6.8
Bug Hunt BenchCode20.527.3+6.8
DeepSWE v1.1Code59.966.7+6.8
GameBench 2Games75.168.7−6.4
ALE-BenchCode19961822−6.2
BrowseComp Extended (Mercor held-out)Search27.521.5−6.0
LMArena Text Arena EloChat & Writing14221456+6.0
GDPval-AA v2.1Professional Work15231658+6.0
MMMU-Pro Extended (Mercor held-out)Multimodal31.926.0−5.9
ARC-AGI-2General QA61.367.1+5.8
AA-PersonalityBench Transparency indexChat & Writing77.271.4−5.8
CharXiv Extended (Mercor held-out)Multimodal82.977.3−5.6
Artificial Analysis Coding Agent Index v1.5Indexes41.447.0+5.5
Artificial Analysis Intelligence Index v5.0.1Indexes36.531.0−5.5
MysteryMechanismScience35.130.6−4.5
Legal Research BenchLegal43.848.1+4.3
SWE-bench VerifiedCode88.984.6−4.3
DAXBench v2Code81.986.2+4.3
GDPval Extended (Mercor held-out rubric score)Professional Work66.061.8−4.2
Immigration BenchLegal61.866.0+4.2
Mystery Game PuzzlesGames30.034.0+4.0
CritPt v2 (agentic, avg@5)Science66.362.3−4.0
SpeciEval (Pooled English Runs)Safety86.990.8+3.9
Conceptual Reasoning IndexIndexes65.362.5−3.8
KernelBench CUDA Normalized (RTX PRO 6000)Code41.437.7−3.7
Excel Modeling BenchmarkFinance & Accounting59.062.7+3.7
SUPERChemScience62.065.7+3.7
Public Benefits Bench v1.1Professional Work63.266.8+3.7
Terminal-Bench 2.1Code85.088.4+3.4
LMArena Code Arena WebDevCode15941618+3.2
MedScribeHealth & Medicine89.586.5−3.0
APEX-1 (Mercor, mean score)Professional Work64.767.7+3.0
GDP.pdf (Artificial Analysis, All-pass)Professional Work20.817.8−3.0
FrontierCode 1.1 Extended - ScoreCode58.461.3+2.9
CritPt (avg@5)Science22.619.7−2.9
τ³-Banking (Mercor run, bm25_grep retrieval)Agents43.045.7+2.7
AA-LCR v1.1Search78.381.0+2.7
LMCAGeneral QA51.148.5−2.6
APEX-SWECode55.157.7+2.6
FrontierCode 1.1 Extended - Pass RateCode64.767.0+2.3
Vals Index v2.1Indexes54.452.1−2.3
DTBenchGeneral QA95.097.3+2.3
LMArena Agent Arena Confirmed SuccessAgents1.6-2.6−2.1
LMArena Agent Arena Steering BurdenAgents2.4-1.8−2.1
ChessBenchGames41.343.3+2.0
MMLU ProGeneral QA87.489.4+2.0
APEX-Accounting 1.1Finance & Accounting62.260.3−1.9
Epoch Capabilities Index (ECI)Indexes160158−1.9
BioMysteryBench (Vals Terminus 2)Science70.472.2+1.9
Artificial Analysis Progress Index v1Indexes326322−1.6
SWE Atlas - Codebase QnACode48.750.3+1.6
LMArena Agent Arena Tool HallucinationAgents-2.11.0+1.5
SciCode 1.0.1 (Artificial Analysis)Code55.056.5+1.5
ARC-AGI-1 Semi-PrivateGeneral QA89.087.5−1.5
ProgramBenchCode2.51.0−1.5
Terminal-Bench Science 0.1 (Vals)Science10.011.4+1.4
MMMU ProMultimodal79.180.5+1.4
AA-LCR Extended (Mercor held-out)Search63.865.2+1.4
Design Arena Overall Code CategoriesCode12801288+1.3
LMArena Agent Arena Bash ReliabilityAgents7.410.0+1.3
LMArena Agent Arena Net Improvement (Steering Burden)Agents0.5-0.2−1.2
CharXiv (descriptive and reasoning combined)Multimodal94.192.9−1.2
Roboflow Vision Evals — Identification, Judged (September)Multimodal86.585.4−1.1
Roboflow Vision Evals — Identification, Strict (September)Multimodal80.281.3+1.1
CVDP (517-problem set)Code57.058.0+1.0
ChessBench AccuracyGames66.967.8+1.0
ChessBench EloGames875891+1.0
Artificial Analysis Intelligence Index v4.3.2Indexes43.444.3+0.9
AA-Briefcase v1.1 EloAgents15231540+0.9
ARC-AGI-3 (Semi-Private)General QA2.82.1−0.7
LiveBench 2.0Agents75.674.9−0.7
LMArena Agent Arena Praise vs ComplaintAgents-3.5-2.5+0.5
BuseyBench SVGMultimodal4.64.6+0.5
GDPval (Mercor 213-task rubric score)Professional Work85.586.0+0.5
Finance Agent v2Finance & Accounting54.153.7−0.5
Vals Science IndexIndexes38.538.1−0.4
Prophet Arena Arena Score (Fixed Context)Forecasting85.485.0−0.4
IOI (Vals v2)Code47.347.6+0.3
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting85.985.6−0.3
GBENCH One-shot Coding GScoreGames55.655.4−0.3
OTIS Mock AIME 2024-2025Math98.999.2+0.3
Humanity's Last Exam (Text-Only)General QA42.742.9+0.2
Blueprint-Bench 2Multimodal0.30.3+0.2
ForecastBench Preliminary Dataset Brier IndexForecasting60.861.0+0.2
AdvancedIFChat & Writing80.880.9+0.1
Code MigrationCode44.644.60.0
AIME 2025Math100.0100.00.0
Harvey's Legal Agent Benchmark - Arbitration International Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Bankruptcy RestructuringLegal0.00.00.0
Harvey's Legal Agent Benchmark - Capital MarketsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Emerging Companies Venture CapitalLegal0.00.00.0
Harvey's Legal Agent Benchmark - Employment LaborLegal0.00.00.0
Harvey's Legal Agent Benchmark - Environmental ESGLegal0.00.00.0
Harvey's Legal Agent Benchmark - ImmigrationLegal0.00.00.0
Harvey's Legal Agent Benchmark - InsuranceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal0.00.00.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.00.00.0
Harvey's Legal Agent Benchmark - TaxLegal0.00.00.0
Harvey's Legal Agent Benchmark - Trusts Estates Private ClientLegal0.00.00.0
Harvey's Legal Agent Benchmark - White Collar Defense InvestigationsLegal0.00.00.0
KernelBench Mega Normalized (RTX PRO 6000)Code0.00.00.0
MedXpertQA MM Extended (Mercor held-out)Health & Medicine48.448.40.0

Claude Haiku 5.5vsGrok 4.6normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 35 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…