Atlas

Compare Models

GPT-5.6 Terra vs Grok 4.7

GPT-5.6 Terra leads by 0.1 capability points (≈95% interval −2.0 to +2.1): statistically tied · P(higher) ≈ 53%

Wins 88 of 186 shared benchmarks (80 lost, 18 tied)

Capability vs. relative cost
GPT-5.6 TerraGrok 4.7Closed 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 4.7capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins88 – 80 · 186 shared, 18 tied

Comparing matching setups only (56 benchmarks excluded) · suites counted once
← GPT-5.6 TerraGrok 4.7 →

GPT-5.6 TerravsGrok 4.7wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 38 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Grok 4.7 minus GPT-5.6 Terra.

FrontierMath Tier 4 (v2) - PrivateMath70.717.1−53.7
ProofBench v1.1Math74.026.0−48.0
Harvey's Legal Agent Benchmark - Capital MarketsLegal0.040.0+40.0
YC-BenchGames6361441666389+36.0
LAB-Bench LitQA2 (SecureBio)Science51.116.9−34.2
Furniture Assembly BenchmarkMultimodal54.220.8−33.3
FrontierMath Tiers 1-3 (v2) - PrivateMath86.053.0−33.0
IOI (Vals v2)Code87.657.7−29.9
SM BenchSafety59.086.3+27.3
LAB-Bench ProtocolQA (SecureBio)Science77.051.4−25.6
Terminal-Bench 2.1 Extended (Mercor held-out)Code36.411.8−24.6
DistressBenchSafety76.452.4−24.0
SWE-bench Verified Extended (Mercor held-out)Code78.755.3−23.4
ARC-AGI-2General QA83.961.4−22.5
APEX-SWECode31.153.4+22.3
CadQueryEval (Corrected Geometry)Code68.088.0+20.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal0.020.0+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 - Intellectual PropertyLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal20.00.0−20.0
Vending-Bench 2Games734310537+19.3
LLM Chess EloGames1293822−19.3
ChartographyMultimodal34.014.7−19.3
GameBench 2Games53.572.3+18.8
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal2.319.5+17.2
Medical Long Context Reasoning (MLCR-AA)Health & Medicine31.715.0−16.7
AA-Briefcase v1.1 EloAgents13321636+16.6
ChessBench CoherenceGames69.786.3+16.6
AA-Omniscience IndexIndexes0.132.0+16.0
Terminal-Bench 4.0 (Artificial Analysis)Code10.125.8+15.7
BuseyBench SVGMultimodal6.14.6−15.1
HPCT (SecureBio)Science68.253.7−14.5
BrowseComp Extended (Mercor held-out)Search25.339.5+14.2
DAYJOB: FinanceFinance & Accounting1.814.5+12.7
LAB-Bench CloningScenarios (SecureBio)Science92.580.1−12.4
CritPt (avg@5)Science30.018.0−12.0
ALE-BenchCode21541823−11.7
Vibe Code Bench v1.1Code74.686.2+11.6
GDPval-AA v2.1Professional Work14561717+11.5
Terminal-Bench 4.0Code26.337.6+11.3
LMArena Agent Arena Net ImprovementAgents0.64.0+11.2
MBCT (SecureBio)Science56.946.1−10.8
Harvey HLAB — Criteria Pass RateLegal79.890.5+10.7
SWE-bench MultilingualCode94.484.2−10.2
BrowseComp (Mercor 130-question subset)Search85.875.8−10.0
Bio Capabilities Index (SecureBio)Indexes155141−9.9
ChessBench EloGames9251076+9.0
ITBench-AAAgents51.042.1−9.0
HardGeoBench <=100miMultimodal36.327.5−8.7
LMArena Agent Arena Net Improvement (Steering Burden)Agents-1.82.6+8.3
DeepsecBench — Revalidated V1Code19.110.9−8.2
DeepsecBench — Revalidated V1 (233 issues)Code19.110.9−8.2
Creative Writing v3 EloChat & Writing18562012+7.9
HardGeoBench <=10miMultimodal25.017.5−7.5
CritPt v2 (agentic, avg@5)Science69.762.3−7.4
GBENCH One-shot Coding GScoreGames63.356.3−7.0
DAYJOB: HealthcareHealth & Medicine1.68.4+6.8
MedXpertQA MM (200-question set)Health & Medicine75.769.0−6.7
ReactBench (39 tasks)Code39.032.3−6.7
MedScribeHealth & Medicine82.989.4+6.5
APEX-AccountingFinance & Accounting40.346.7+6.4
ARC-AGI-1 Semi-PrivateGeneral QA96.590.2−6.3
SAGEMath47.040.8−6.2
AA-LCR Extended (Mercor held-out)Search74.468.2−6.2
DeepSWE v1.1Code64.871.0+6.2
τ³-Banking (Mercor run, bm25_grep retrieval)Agents38.832.6−6.2
LMArena Agent Arena Confirmed SuccessAgents-7.25.1+6.1
MedCodeHealth & Medicine43.449.6+6.1
Mystery Game PuzzlesGames35.029.0−6.0
Chess Puzzles (task version 2.0)Games47.041.0−6.0
Ruby LLM Benchmarks — Program FixerCode74.068.0−6.0
AutomationBench-AA 1.0.6Agents59.665.6+5.9
Legal Research BenchLegal41.347.1+5.8
SRE Bench — Fully SolvedCode9.53.8−5.7
LMCAGeneral QA55.049.4−5.6
Finance Agent v2 - Financial ModelingFinance & Accounting23.017.7−5.4
GPQA DiamondGeneral QA87.492.7+5.4
Design Arena Overall Code CategoriesCode12381270+5.3
Humanity's Last Exam DiamondGeneral QA28.823.6−5.2
CursorBench 4.0Code41.346.3+5.0
Longform Creative WritingChat & Writing78.083.0+5.0
Finance Agent v2 - ComparablesFinance & Accounting47.042.2−4.8
FrontierCode 2.0 (Artificial Analysis)Code23.928.7+4.8
LMArena Text Arena EloChat & Writing14681442−4.5
Vision Arena OverallMultimodal12661249−4.5
Artificial Analysis Intelligence Index v4.3.2Indexes42.146.4+4.4
LMArena Agent Arena Bash ReliabilityAgents0.79.3+4.3
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting72.768.6−4.1
APEX-1 (Mercor, mean score)Professional Work69.565.4−4.1
CVDP (517-problem set)Code50.954.9+4.0
Epoch Capabilities Index (ECI)Indexes161156−3.9
Tax Agent BenchFinance & Accounting65.269.0+3.8
Conceptual Reasoning IndexIndexes65.763.1−3.5
HardGeoBench Avg DistanceMultimodal19912104−3.5
Finance Agent v2 - AdjustmentsFinance & Accounting49.746.3−3.5
Olam Social Poker EloGames15111508−3.3
Public Benefits Bench v1.1Professional Work62.465.6+3.2
LMArena Agent Arena Bash RecoveryAgents0.97.2+3.2
Artificial Analysis Intelligence Index v5.0.1Indexes32.335.5+3.1
Concept Synth INDUCTION Challenge64 Round 1General QA37.534.4−3.1
APEX-Agents 1.1Professional Work71.868.7−3.1
Financial Coaching BenchFinance & Accounting71.074.1+3.1
Terminal-Bench 2.1Code88.085.0−3.0
Code MigrationCode47.844.8−3.0
CharXiv Extended (Mercor held-out)Multimodal81.178.2−2.9
GPQA-Bio (SecureBio)Science85.382.5−2.8
SciCode 1.0.1 (Artificial Analysis)Code55.057.8+2.8
Finance Agent v2 - Earnings AnalysisFinance & Accounting77.374.6−2.7
DTBenchGeneral QA93.396.0+2.7
CyberBench v1.1 OverallCybersecurity72.169.5−2.6
Roboflow Vision Evals (October 2026, five tasks)Multimodal66.063.5−2.5
LMArena Agent Arena Praise vs ComplaintAgents-1.93.1+2.5
Finance Agent v2 - Disclosure AnalysisFinance & Accounting66.964.5−2.5
Terminal-Bench-Science 0.1Science13.811.4−2.4
ChessBench AccuracyGames71.974.0+2.3
WCB (SecureBio)Science59.857.6−2.2
AdvancedIFChat & Writing78.180.2+2.1
Toloka Arena (v1)Agents41.439.3−2.1
CharXiv (descriptive and reasoning combined)Multimodal94.192.0−2.1
GeneBench-Pro Public Release SubsetScience8.010.0+2.0
Roboflow Vision Evals (September 2026 protocol)Multimodal73.871.9−1.9
DeepsecBench-AA (Artificial Analysis)Cybersecurity25.126.9+1.9
Vals Index v2.1Indexes53.154.9+1.9
LMArena Agent Arena SteerabilityAgents4.10.5−1.8
DeepResearch Bench II (Mercor 130-task set)Search47.249.0+1.8
SWE Atlas - Codebase QnACode45.247.0+1.8
RedlineBenchLegal47.849.5+1.7
Blueprint-Bench 2Multimodal0.30.3+1.7
SpeciEval (Refusal-Aware Retries, Historical)Safety98.196.4−1.7
SAGE v1.1Math87.088.7+1.7
SpeciEval (Pooled English Runs)Safety98.196.4−1.7
OTIS Mock AIME 2024-2025Math99.798.1−1.7
Extended NYT Connections 940 (Quadratic v1)Games78.476.8−1.6
Terminal-Bench Science 0.1 (Vals)Science10.011.4+1.4
GDP.pdf (Artificial Analysis, All-pass)Professional Work24.623.2−1.4
ObviousBenchGeneral QA97.999.3+1.4
MMMU ProMultimodal79.578.2−1.3
SWE-bench VerifiedCode77.879.1+1.3
Humanity's Last Exam (Text-Only)General QA41.943.1+1.3
HardGeoBench Country AccuracyMultimodal45.043.8−1.2
Finance Agent v2 - Market AnalysisFinance & Accounting60.661.8+1.2
PubMedQA (SecureBio)Health & Medicine80.479.2−1.2
ARC-AGI-3 (Semi-Private)General QA0.71.8+1.2
Finance Agent v2 - PrecedentsFinance & Accounting24.825.9+1.1
AA-LCR v1.1Search79.080.0+1.0
Concept Synth INDUCTION Challenge100 Round 1General QA25.024.0−1.0
Eval Connections One-Shot with TrapsGames86.085.0−1.0
MedXpertQA MM Extended (Mercor held-out)Health & Medicine51.052.0+1.0
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting67.868.7+0.9
WMDP-Bio (SecureBio)Science89.788.8−0.9
LiveBench 2.0Agents75.074.1−0.9
LMArena Agent Arena Steering BurdenAgents-0.5-2.3−0.9
Excel Modeling BenchmarkFinance & Accounting66.267.0+0.8
LegalBenchLegal85.184.4−0.7
MMLU ProGeneral QA88.188.8+0.7
SWE Atlas Codebase QnA Extended (Mercor held-out)Code16.015.3−0.7
Vals Index v2Indexes59.660.2+0.6
ACCoRDGeneral QA74.774.0−0.6
Finance Agent v2 - All PassFinance & Accounting42.842.1−0.6
Artificial Analysis Progress Index v1Indexes324323−0.4
Creative Writing v3 Rubric ScoreChat & Writing15.915.9+0.3
MMLU-Bio (SecureBio)Science90.790.9+0.2
VCT v2 (SecureBio)Science41.942.0+0.1
LMArena Agent Arena Tool HallucinationAgents1.01.00.0
AIME 2025Math100.0100.00.0
Harvey's Legal Agent Benchmark - Arbitration International Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Banking FinanceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Bankruptcy RestructuringLegal0.00.00.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.00.00.0
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 - 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
ProgramBenchCode0.50.50.0
Senior SWE-BenchCode27.427.40.0

GPT-5.6 TerravsGrok 4.7normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 38 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…