Atlas

Compare Models

GPT-5.6 Terra vs Grok 4.6

GPT-5.6 Terra leads by 0.7 capability points (≈95% interval −1.3 to +2.7): statistically tied · P(higher) ≈ 75%

Wins 105 of 220 shared benchmarks (98 lost, 17 tied)

Capability vs. relative cost
GPT-5.6 TerraGrok 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.

GPT-5.6 TerravsGrok 4.6capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins105 – 98 · 220 shared, 17 tied

Comparing matching setups only (65 benchmarks excluded) · suites counted once
← GPT-5.6 TerraGrok 4.6 →

GPT-5.6 TerravsGrok 4.6wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 48 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 GPT-5.6 Terra.

IOI (Vals v2)Code87.647.6−40.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal0.040.0+40.0
FrontierMath Tier 4 (v2) - PrivateMath70.731.7−39.0
LLM Chess EloGames1293443−34.9
YC-BenchGames6361441568241+32.5
SM BenchSafety59.089.0+30.0
BioSecBench-RefusalSafety34.364.2+29.9
MRCR v2 (8-needle, Context Arena) - 128K pointwiseGeneral QA80.252.6−27.5
APEX-SWECode31.157.7+26.6
Document ArenaProfessional Work14681438−26.4
HLE Extended (Mercor held-out)General QA70.745.8−24.9
ProofBench v1.1Math74.051.0−23.0
LAB-Bench LitQA2 (SecureBio)Science51.128.9−22.2
MMMU-Pro Extended (Mercor held-out)Multimodal47.426.0−21.4
TxBench Antibody Discovery V0Science31.051.9+20.9
FrontierMath Tiers 1-3 (v2) - PrivateMath86.066.0−20.0
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
Medical Long Context Reasoning (MLCR-AA)Health & Medicine31.712.2−19.4
Wiz Cyber Model Arena v3.9Cybersecurity64.445.5−18.9
SpatialBench-Long V0 VerifiedScience37.919.7−18.2
SAGEMath47.028.9−18.1
ARC-AGI-2General QA83.967.1−16.8
GameBench 2Games53.568.7+15.2
AA-Omniscience IndexIndexes0.130.5+15.2
MRCR v2 (8-needle, Context Arena) - 128K AUCGeneral QA89.474.4−15.0
BuseyBench SVGMultimodal6.14.6−14.7
GDPval Extended (Mercor held-out rubric score)Professional Work47.261.8+14.6
Gray Swan IPI Attack Success (k=15)Safety37.351.8−14.5
Furniture Assembly BenchmarkMultimodal54.240.0−14.2
Chess Puzzles (task version 2.0)Games47.033.0−14.0
ChartographyMultimodal34.020.5−13.5
Olam Social Poker EloGames15111499−13.0
Bio Capabilities Index (SecureBio)Indexes155137−12.9
Harvey HLAB — Criteria Pass RateLegal79.892.5+12.8
Hedge-BenchFinance & Accounting13.826.4+12.7
AA-Briefcase EloAgents13371545+12.1
ALE-BenchCode21541822−11.8
DAYJOB: FinanceFinance & Accounting1.812.8+11.0
WeirdML (v2)Science78.367.3−11.0
SkillsBench without SkillsAgents48.838.2−10.7
TxBench Oligo Discovery V0Science33.344.0+10.6
ScBench-Long V0 VerifiedScience34.924.2−10.6
𝜏³-BankingAgents40.250.7+10.5
Vending-Bench 2Games73439047+10.3
CritPt (avg@5)Science30.019.7−10.3
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting67.877.8+10.0
EnterpriseOps-Gym-AAAgents38.548.3+9.8
WCB (SecureBio)Science59.850.3−9.5
AA-LCR Extended (Mercor held-out)Search74.465.2−9.2
Terminal-Bench 2.1 Extended (Mercor held-out)Code36.445.5+9.1
GDPval-AA v2Professional Work14791663+9.1
ARC-AGI-1 Semi-PrivateGeneral QA96.587.5−9.0
Shroombench (mushroom identification, top-1)Multimodal35.044.0+9.0
GDPval-AA v2.1Professional Work14561658+8.9
LAB-Bench ProtocolQA (SecureBio)Science77.068.1−8.9
HLE-Bio (SecureBio)Science38.029.2−8.8
BioSecBench-Function V0Science33.542.0+8.5
AHK-EvalCode88.997.2+8.3
Toloka Arena (v1)Agents41.449.7+8.3
Design Arena Overall Code CategoriesCode12381288+8.3
BabaIsBench Stage 0: The IntroGames88.096.0+8.0
GBENCH One-shot Coding GScoreGames63.355.4−8.0
Terminal-Bench 2.1Code80.188.0+7.9
Terminal-Bench-Science 0.1Science13.86.2−7.6
CritPt v2 (agentic, avg@5)Science69.762.3−7.4
AutomationBench-AA 1.0.6Agents59.667.0+7.3
Humanity's Last Exam DiamondGeneral QA28.821.5−7.3
SpeciEval (Refusal-Aware Retries, Historical)Safety98.190.8−7.2
SpeciEval (Pooled English Runs)Safety98.190.8−7.2
CVDP (517-problem set)Code50.958.0+7.1
Terminal-Bench 4.0 (Artificial Analysis)Code10.117.2+7.1
GeneBench-Pro Public Release SubsetScience8.015.0+7.0
Roboflow Vision Evals (October 2026, five tasks)Multimodal66.059.1−6.9
GDP.pdf (Artificial Analysis, All-pass)Professional Work24.617.8−6.8
Legal Research BenchLegal41.348.1+6.7
HPCT (SecureBio)Science68.261.6−6.6
DeepSWE v1.1Code60.266.7+6.6
LMCAGeneral QA55.048.5−6.5
MBCT (SecureBio)Science56.950.4−6.5
Tax Agent BenchFinance & Accounting65.271.4+6.2
HealthBench Professional (Length-Adjusted)Health & Medicine57.751.7−6.0
Finance Agent v2 - ComparablesFinance & Accounting47.041.2−5.8
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science5.811.5+5.8
HardGeoBench Avg DistanceMultimodal19911807+5.7
EpiBenchScience39.934.3−5.7
GPQA DiamondGeneral QA89.995.5+5.6
NoCoT-Bench NCKI (kspine v3)General QA104112+5.5
APEX-Agents 1.1Professional Work71.877.3+5.5
GDPval (Mercor 213-task rubric score)Professional Work80.786.0+5.3
ReactBench (39 tasks)Code39.033.8−5.1
Roboflow Vision Evals (September 2026 protocol)Multimodal73.868.7−5.1
SWE Atlas - Codebase QnACode45.250.3+5.1
TaxEval v2Finance & Accounting76.271.1−5.1
CursorBench 3.2Code64.969.9+5.0
Eval Connections One-Shot with TrapsGames86.091.0+5.0
HardGeoBench <=10miMultimodal25.020.0−5.0
LMArena Agent Arena Bash ReliabilityAgents0.710.0+4.6
Public Benefits Bench v1.1Professional Work62.466.8+4.5
CyberBench v1.1 OverallCybersecurity72.167.7−4.4
ChessBench AccuracyGames71.967.8−4.4
Conceptual Reasoning IndexIndexes65.762.5−4.4
FrontierCode 2.0 (Artificial Analysis)Code23.919.8−4.1
Finance Agent v2 - Financial ModelingFinance & Accounting23.018.9−4.1
CadQueryEval (Corrected Geometry)Code68.072.0+4.0
DTBenchGeneral QA93.397.3+4.0
BrowseComp Extended (Mercor held-out)Search25.321.5−3.8
CharXiv Extended (Mercor held-out)Multimodal81.177.3−3.8
HardGeoBench <=100miMultimodal36.332.5−3.7
MedScribeHealth & Medicine82.986.5+3.7
Excel Modeling BenchmarkFinance & Accounting66.262.7−3.5
Finance Agent v2 - Earnings AnalysisFinance & Accounting77.373.9−3.4
SWE Atlas Codebase QnA Extended (Mercor held-out)Code16.019.3+3.3
Code MigrationCode47.844.6−3.2
ZeroBench Main Mean pass@1 (five samples)Multimodal14.211.0−3.2
MortgageTax (Vals all 1,258 documents)Finance & Accounting67.364.2−3.1
LMArena Agent Arena Net Improvement (Steering Burden)Agents-1.8-0.2+3.1
SkillsBench with SkillsAgents58.955.8−3.1
ACCoRDGeneral QA74.771.6−3.1
ZeroBench Main Questions pass^5Multimodal10.07.0−3.0
DeepsecBench — Revalidated V1Code19.116.2−3.0
DeepsecBench — Revalidated V1 (233 issues)Code19.116.1−3.0
Finance Agent v2 - PrecedentsFinance & Accounting24.821.9−2.9
AdvancedIFChat & Writing78.180.9+2.8
MedXpertQA MM Extended (Mercor held-out)Health & Medicine51.048.4−2.6
Finance Agent v2 - Disclosure AnalysisFinance & Accounting66.964.4−2.6
Blueprint-Bench 2Multimodal0.30.3+2.4
LMArena Agent Arena Bash RecoveryAgents0.95.7+2.4
Epoch Capabilities Index (ECI)Indexes161158−2.4
LMArena Agent Arena Confirmed SuccessAgents-7.2-2.6+2.3
LiveCodeBench (Vals AI public dataset)Code85.988.2+2.3
Artificial Analysis Intelligence Index v4.3.2Indexes42.144.3+2.2
Artificial Analysis Intelligence Index v4.3Indexes42.344.4+2.2
GPQA-Bio (SecureBio)Science85.387.4+2.1
ObviousBenchGeneral QA97.9100.0+2.1
ChessBench EloGames925891−2.0
AA-LCR v1.1Search79.081.0+2.0
WordleBenchGames91.089.0−2.0
ZeroBench (Main questions, pass@5)Multimodal19.017.0−2.0
LMArena Text Arena EloChat & Writing14681456−2.0
APEX-1 (Mercor, mean score)Professional Work69.567.7−1.8
BrowseComp (Mercor 130-question subset)Search85.884.0−1.8
LAB-Bench CloningScenarios (SecureBio)Science92.594.2+1.7
SWE-bench MultilingualCode94.492.7−1.7
Vibe Code Bench v1.1Code74.676.2+1.6
Finance Agent v2 - Market AnalysisFinance & Accounting60.662.2+1.6
LMArena Agent Arena Net ImprovementAgents0.61.1+1.6
Extended NYT Connections 940 (Quadratic v1)Games78.480.0+1.6
HackerRank-BenchCode26.127.7+1.6
Concept Synth INDUCTION Challenge64 Round 1General QA37.535.9−1.6
SciCode 1.0.1 (Artificial Analysis)Code55.056.5+1.5
ARC-AGI-3 (Semi-Private)General QA0.72.1+1.5
Terminal-Bench Science 0.1 (Vals)Science10.011.4+1.4
Artificial Analysis Intelligence Index v5.0.1Indexes32.331.0−1.3
MMLU ProGeneral QA88.189.4+1.3
MedCodeHealth & Medicine43.444.7+1.3
HardGeoBench Country AccuracyMultimodal45.043.8−1.2
SAB ResearchFinance & Accounting93.494.6+1.2
NoCoT-Bench NCRI 15.2General QA89.086.8−1.2
LegalBenchLegal85.186.3+1.2
SciCode (main problems)Science15.814.6−1.2
DAYJOB: HealthcareHealth & Medicine1.62.8+1.2
CharXiv (descriptive and reasoning combined)Multimodal94.192.9−1.2
MMLU-Bio (SecureBio)Science90.791.8+1.1
Senior SWE-BenchCode27.426.3−1.1
VCT (SecureBio)Science47.248.3+1.1
MMMU ProMultimodal79.580.5+1.0
Humanity's Last Exam (Text-Only)General QA41.942.9+1.0
Mystery Game PuzzlesGames35.034.0−1.0
Vals Index v2.1Indexes53.152.1−1.0
Terminal-Bench 4.0Code26.325.3−1.0
CorpFin v2Finance & Accounting65.366.2+0.9
Artificial Analysis Progress Index v1Indexes324322−0.8
Vision Arena OverallMultimodal12661263−0.8
SUPERChemScience64.965.7+0.8
BioSecBench-SurveillanceScience51.051.7+0.7
RedlineBenchLegal47.848.5+0.7
SWE-bench Verified Extended (Mercor held-out)Code78.778.0−0.7
LMArena Agent Arena Steering BurdenAgents-0.5-1.8−0.6
VariantBenchScience38.738.1−0.6
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting72.773.2+0.6
SpatialBench V0 VerifiedScience70.469.9−0.6
OTIS Mock AIME 2024-2025Math99.799.2−0.6
APEX-Accounting 1.1Finance & Accounting59.860.3+0.5
ChessBench CoherenceGames69.769.2−0.5
MedXpertQA MM (200-question set)Health & Medicine75.775.2−0.5
ProgramBenchCode0.51.0+0.5
LMArena Agent Arena SteerabilityAgents4.13.3−0.4
Vals Index v2Indexes59.659.2−0.4
LMArena Agent Arena Praise vs ComplaintAgents-1.9-2.5−0.3
HieroglyphBenchMultimodal20.520.3−0.2
SWE-bench VerifiedCode95.495.6+0.2
Finance Agent v2 - All PassFinance & Accounting42.842.9+0.1
CursorBench 4.0Code41.341.4+0.1
Finance Agent v2 - AdjustmentsFinance & Accounting49.749.6−0.1
LiveBench 2.0Agents75.074.90.0
LMArena Agent Arena Tool HallucinationAgents1.01.00.0
AIME 2025Math100.0100.00.0
Concept Synth INDUCTION Challenge100 Round 1General QA25.025.00.0
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 - Structured Finance SecuritizationLegal20.020.00.0
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

GPT-5.6 TerravsGrok 4.6normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 48 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…