Atlas

Compare Models

Gemini 4 Argon vs GPT-5.6 Terra

Gemini 4 Argon leads by 5.2 capability points (≈95% interval +3.0 to +7.5): a clear lead over GPT-5.6 Terra · P(higher) ≈ 100%

Wins 154 of 180 shared benchmarks (14 lost, 12 tied)

Capability vs. relative cost
Gemini 4 ArgonGPT-5.6 TerraClosed 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.

Gemini 4 ArgonvsGPT-5.6 Terracapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins154 – 14 · 180 shared, 12 tied

Comparing matching setups only (56 benchmarks excluded) · suites counted once
← Gemini 4 ArgonGPT-5.6 Terra →

Gemini 4 ArgonvsGPT-5.6 Terrawins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Center for AI Safety and 37 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-5.6 Terra minus Gemini 4 Argon.

Terminal-Bench 4.0 (Artificial Analysis)Code55.11.0−54.0
ARC-AGI-3 (Semi-Private)General QA49.80.0−49.8
HieroglyphBenchMultimodal69.720.5−49.2
BrowseComp Extended (Mercor held-out)Search72.025.3−46.7
Agent Red Teaming k=100 (CAIS)Safety10.254.6−44.3
EnigmaEvalGeneral QA64.623.1−41.5
Harvey's Legal Agent Benchmark - Capital MarketsLegal40.00.0−40.0
Harvey's Legal Agent Benchmark - Emerging Companies Venture CapitalLegal40.00.0−40.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal40.00.0−40.0
Harvey's Legal Agent Benchmark - Trusts Estates Private ClientLegal40.00.0−40.0
HardGeoBench Avg DistanceMultimodal7411991−38.6
ChartographyMultimodal71.634.0−37.6
TextQuests Game ProgressGames77.340.5−36.8
GameBench 2Games89.253.5−35.7
AutomationBench 1.0.6 (Zapier)Agents50.114.8−35.3
Terminal-Bench-Science 0.1Science48.613.8−34.8
SRE Bench — Fully SolvedCode44.39.5−34.7
Terminal-Bench Science 0.1 (Vals)Science44.310.0−34.3
NoCoT-Bench NCRI 15.2General QA15189.0−34.2
MindCube TinyMultimodal96.762.6−34.1
ChessBench EloGames1485925−33.3
Humanity's Sixth SenseMultimodal55.422.7−32.7
MASKSafety61.093.3+32.3
Concept Synth INDUCTION Challenge100 Round 1General QA57.025.0−32.0
Terminal-Bench 4.0Code57.426.3−31.1
Terminal-Bench 2.1Code57.788.0+30.3
GDPval Extended (Mercor held-out rubric score)Professional Work76.447.2−29.2
SpatialViz-BenchMultimodal94.265.2−29.1
Mystery Game PuzzlesGames64.035.0−29.0
Last Translation Benchmark (LTBv1-eval, blind)Chat & Writing40.912.2−28.8
LisanBench Difficulty-WeightedGames38491197−28.4
Concept Synth INDUCTION Challenge64 Round 1General QA65.637.5−28.1
SEC-Bench Pro - release-reported cohortCybersecurity83.457.7−25.7
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal27.42.3−25.1
ProofBench v1.1Math99.074.0−25.0
Finance Agent v2 - PrecedentsFinance & Accounting49.824.8−25.0
LMArena Agent Arena Net Improvement (Steering Burden)Agents10.7-1.8−23.8
Blueprint-Bench 2Multimodal0.50.3−23.6
ALE-BenchCode28052154−23.0
ERQAMultimodal86.563.6−22.9
ChessBench CoherenceGames92.669.7−22.9
HardGeoBench <=100miMultimodal58.836.3−22.5
BioTIER RefusalSafety87.265.6−21.6
AA-Omniscience IndexIndexes42.40.1−21.1
Code MigrationCode68.247.8−20.4
HardGeoBench <=10miMultimodal45.025.0−20.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - ImmigrationLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal0.020.0+20.0
Humanity's Last ExamGeneral QA55.635.9−19.7
τ³-Banking (Mercor run, bm25_grep retrieval)Agents58.138.8−19.3
HackerRank-BenchCode45.426.1−19.3
Finance Agent v2 - Market AnalysisFinance & Accounting79.760.6−19.1
LAB-Bench ProtocolQA (SecureBio)Science95.977.0−18.9
HardGeoBench Country AccuracyMultimodal63.845.0−18.7
Furniture Assembly BenchmarkMultimodal72.554.2−18.3
APEX-SWECode49.231.1−18.1
Artificial Analysis Intelligence Index v5.0.1Indexes50.432.3−18.1
AutomationBench-AA 1.0.6Agents77.559.6−17.9
NoCoT-Bench NCKI (kspine v3)General QA128104−17.1
DAYJOB: FinanceFinance & Accounting18.81.8−17.0
SpeciEval (Pooled English Runs)Safety81.398.1+16.8
Wiz Cyber Model Arena v3.9Cybersecurity77.260.9−16.3
MedXpertQA MM Extended (Mercor held-out)Health & Medicine66.751.0−15.7
APEX-Agents 1.1Professional Work87.471.8−15.6
AA-Briefcase v1.1 EloAgents14851202−15.5
VCT v2 (SecureBio)Science57.341.9−15.4
MedCodeHealth & Medicine58.843.4−15.4
BuseyBench SVGMultimodal7.66.1−14.8
Design Arena Overall Code CategoriesCode13261238−14.6
Harvey HLAB — Criteria Pass RateLegal94.079.8−14.3
Machiavelli Deception (CAIS)Safety85.080.4+14.0
Tax Agent BenchFinance & Accounting79.065.2−13.8
Humanity's Last Exam DiamondGeneral QA42.528.8−13.7
Bio Capabilities Index (SecureBio)Indexes174155−13.6
Legal Research BenchLegal54.841.3−13.5
IntPhys 2Multimodal74.661.3−13.4
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine13.927.2+13.3
LisanBench Path LengthGames75603303−13.2
TextQuests Average Cumulative HarmSafety26.122.2+13.2
DeepSWE v1.1Code77.964.8−13.1
Finance Agent v2 - All PassFinance & Accounting55.842.8−13.0
LMArena Agent Arena Praise vs ComplaintAgents24.1-1.9−13.0
LLM Chess EloGames16091293−13.0
LVBenchMultimodal91.778.9−12.8
IOI (Vals v2)Code100.087.6−12.4
BullshitBench v2 LeaderboardSafety66.054.0−12.0
Chess Puzzles (task version 2.0)Games59.047.0−12.0
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science17.65.8−11.8
APEX-Accounting 1.1Finance & Accounting71.559.8−11.7
CursorBench 4.0Code52.941.3−11.6
VCT (SecureBio)Science58.447.2−11.2
LMArena Agent Arena Bash RecoveryAgents-21.40.9+11.1
Finance Agent v2 - AdjustmentsFinance & Accounting60.549.7−10.8
GDPval-AA v2.1Professional Work16341388−10.8
ChessBench AccuracyGames81.871.9−10.6
Artificial Analysis Intelligence Index v4.3.2Indexes52.642.1−10.5
LMArena Agent Arena Confirmed SuccessAgents13.2-7.2−10.2
RiemannBenchMath76.066.4−9.6
LMArena Text Arena EloChat & Writing15221468−9.5
Conceptual Reasoning IndexIndexes72.465.7−9.3
DAYJOB: HealthcareHealth & Medicine10.81.6−9.2
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting77.067.8−9.2
LiveBench 2.0Agents84.175.0−9.1
Excel Modeling BenchmarkFinance & Accounting75.266.2−9.0
GPQA DiamondGeneral QA96.487.4−9.0
MMMU ProMultimodal88.379.5−8.8
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting81.572.7−8.8
LMCAGeneral QA63.755.0−8.7
DistressBenchSafety68.676.4+7.8
LABBench2Science88.881.2−7.6
Finance Agent v2 - Disclosure AnalysisFinance & Accounting74.566.9−7.6
Finance Agent v2 - Earnings AnalysisFinance & Accounting84.877.3−7.5
LAB-Bench CloningScenarios (SecureBio)Science100.092.5−7.5
PCT Political Inconsistency (CAIS)Safety38.746.2−7.5
Public Benefits Bench v1.1Professional Work69.862.4−7.4
ARC-AGI-2General QA91.383.9−7.3
FrontierMath Tier 4 (v2) - PrivateMath78.070.7−7.3
SWE-Bench Pro V2 — FullCode99.292.4−6.8
LMArena Agent Arena Steering BurdenAgents13.2-0.5−6.8
SciCode 1.0.1 (Artificial Analysis)Code61.855.0−6.8
Finance Agent v2 - Financial ModelingFinance & Accounting29.823.0−6.8
SAGEMath53.647.0−6.6
Vibe Code Bench 1–100Code21.214.8−6.4
MBCT (SecureBio)Science63.256.9−6.3
Artificial Analysis Progress Index v1Indexes342324−6.3
Eval Connections One-Shot with TrapsGames92.086.0−6.0
HPCT (SecureBio)Science74.268.2−6.0
CyberBench v1.1 OverallCybersecurity77.972.1−5.8
GDPval (Mercor 213-task rubric score)Professional Work85.980.7−5.2
Creative Writing v3 EloChat & Writing17581856+5.0
Epoch Capabilities Index (ECI)Indexes167161−4.6
MedScribeHealth & Medicine87.482.9−4.6
GBENCH One-shot Coding GScoreGames67.963.3−4.5
RCM-Bench-1Health & Medicine91.587.0−4.5
DTBenchGeneral QA97.693.3−4.3
GPQA-Bio (SecureBio)Science89.385.3−4.0
HPCT Refusal (CAIS)Safety29.025.0−4.0
Finance Agent v2 - ComparablesFinance & Accounting51.047.0−3.9
LegalBenchLegal88.385.1−3.2
CritPt v2 (agentic, avg@5)Science72.969.7−3.1
LMArena Agent Arena Bash ReliabilityAgents-5.40.7+3.1
Longform Creative WritingChat & Writing80.978.0−2.9
CritPt (avg@5)Science27.130.0+2.9
ACCoRDGeneral QA77.374.7−2.6
SAGE v1.1Math89.587.0−2.4
Financial Coaching BenchFinance & Accounting73.171.0−2.2
ARC-AGI-1 Semi-PrivateGeneral QA98.596.5−2.0
ProgramBenchCode2.50.5−2.0
PubMedQA (SecureBio)Health & Medicine82.480.4−2.0
FrontierCode 2.0 (Artificial Analysis)Code25.723.9−1.8
WMDP-Bio (SecureBio)Science91.589.7−1.8
Creative Writing v3 Rubric ScoreChat & Writing15.615.9+1.7
MMLU-Bio (SecureBio)Science92.190.7−1.4
WCB (SecureBio)Science61.159.8−1.3
GDP.pdf (Artificial Analysis, All-pass)Professional Work25.624.6−1.0
AA-LCR v1.1Search79.779.0−0.7
Senior SWE-BenchCode27.727.4−0.3
AA-LCR Extended (Mercor held-out)Search74.774.4−0.3
OTIS Mock AIME 2024-2025Math100.099.7−0.3
LMArena Agent Arena Tool HallucinationAgents0.51.0+0.2
FrontierMath Tiers 1-3 (v2) - PrivateMath86.086.00.0
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 - Employment LaborLegal0.00.00.0
Harvey's Legal Agent Benchmark - Environmental ESGLegal0.00.00.0
Harvey's Legal Agent Benchmark - InsuranceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.00.00.0
Harvey's Legal Agent Benchmark - TaxLegal0.00.00.0
Harvey's Legal Agent Benchmark - White Collar Defense InvestigationsLegal0.00.00.0
VCT-Refusal (CAIS)Safety10.710.70.0

Gemini 4 ArgonvsGPT-5.6 Terranormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Center for AI Safety and 37 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…