Atlas

Compare Models

DeepSeek V4.1 Flash vs Grok 4.6

Grok 4.6 leads by 0.6 capability points (≈95% interval −1.3 to +2.5): statistically tied · P(higher) ≈ 73%

Wins 121 of 222 shared benchmarks (81 lost, 20 tied)

Capability vs. relative cost
DeepSeek V4.1 FlashGrok 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.

DeepSeek V4.1 FlashvsGrok 4.6capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins81 – 121 · 222 shared, 20 tied

Comparing matching setups only (11 benchmarks excluded) · suites counted once
← DeepSeek V4.1 FlashGrok 4.6 →

DeepSeek V4.1 FlashvsGrok 4.6wins by categoryData Nov 15, 2026Data: Vals AI, benchmarks.bio, Arena and 42 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 DeepSeek V4.1 Flash.

AA-PersonalityBench Organization indexChat & Writing10014.3−86.5
KernelBench Mega Normalized (RTX PRO 6000)Code48.20.0−48.2
BioSecBench-Refusal V1 Clean RerunsSafety11.346.9+35.6
AA-PersonalityBench Informativeness indexChat & Writing79.144.8−34.6
LLM Creative Story WritingChat & Writing-0.5-2.9−27.0
APEX-Agents 1.1Professional Work50.577.3+26.8
LLM Chess EloGames1087443−26.4
EnterpriseBench: CoreCraftProfessional Work41.565.6+24.1
SkillsBench without SkillsAgents61.738.2−23.5
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 - Energy Natural ResourcesLegal20.040.0+20.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.020.0+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's Legal Agent Benchmark - Trusts Estates Private ClientLegal20.00.0−20.0
SAGEMath47.928.9−19.0
Concept Synth INDUCTION Challenge64 Round 1General QA17.235.9+18.8
Eval Connections One-Shot with TrapsGames73.091.0+18.0
AA-Omniscience IndexIndexes-5.330.5+17.9
CloudOpsBench v1.0Agents21.839.7+17.9
SWE Atlas Codebase QnA Extended (Mercor held-out)Code36.719.3−17.4
Harvey LAB held-out set (Fireworks run)Legal5.722.1+16.4
SpatialBench-Long V0 VerifiedScience34.919.7−15.2
Chess Puzzles (task version 2.0)Games48.033.0−15.0
TxBench Preclinical Pharmacology V1 Clean RerunsScience46.761.7+15.0
τ³-Banking (Fireworks, banking_knowledge)Agents35.750.7+14.9
SkillsBench with SkillsAgents69.855.8−14.0
Concept Synth INDUCTION Challenge100 Round 1General QA11.025.0+14.0
TxBench Antibody Discovery V1 (100 evaluations)Science39.753.0+13.3
ComplexConstraintsChat & Writing49.736.5−13.2
TxBench Antibody Discovery V0Science39.051.9+12.8
BioSecBench-Structure V1 (58 evaluations)Science27.615.5−12.1
FoodTruck BenchGames5017959537+12.0
ORCA-bench Hard RCA (Fireworks 324-task subset)Agents20.332.1+11.8
MedXpertQA MM (200-question set)Health & Medicine63.575.2+11.7
TxBench-Antibody-Development V1 (60 evaluations)Science14.426.1+11.7
AA-PersonalityBench Transparency indexChat & Writing60.071.4+11.5
AA-PersonalityBench Imagination indexChat & Writing31.520.0−11.5
TxBench Oligo Discovery V1 (113 evaluations)Science34.846.0+11.2
DAXBench v2Code75.586.2+10.6
Medical Long Context Reasoning (MLCR-AA)Health & Medicine22.812.2−10.6
APEX-Accounting 1.1Finance & Accounting50.160.3+10.2
TxBench Oligo Discovery V0Science33.944.0+10.0
Harvey LAB: In-House Contracting (held-out)Legal0.310.0+9.7
AA-PersonalityBench Warmth indexChat & Writing41.031.5−9.6
Extended NYT Connections 940 (Quadratic v1)Games89.680.0−9.6
MysteryMechanismScience21.230.6+9.5
VariantBench V1 Clean RerunsScience31.140.4+9.3
KernelBench CUDA Normalized (RTX PRO 6000)Code47.037.7−9.3
RuneBenchGames5.76.3+9.3
KernelBench Hard Normalized (RTX PRO 6000)Code69.760.5−9.3
SimpleBenchGeneral QA66.775.9+9.2
Mystery Game PuzzlesGames43.034.0−9.0
SUPERChemScience56.765.7+9.0
LMArena Agent Arena Net ImprovementAgents3.91.1−9.0
Tax Agent BenchFinance & Accounting62.571.4+8.9
AA-Briefcase EloAgents14241577+8.9
ChartographyMultimodal11.620.5+8.9
BlenderWorldCode65.757.0−8.7
Vibe Code Bench v1.1Code84.776.2−8.5
Vals Science IndexIndexes29.638.1+8.4
CharXiv Extended (Mercor held-out)Multimodal68.977.3+8.4
τ³-Banking (Mercor run, bm25_grep retrieval)Agents37.845.7+7.9
VIALSMultimodal12.620.5+7.9
EpiBenchScience42.134.3−7.9
ReactBench (39 tasks): Fixing ReactCode18.826.3+7.5
SWEeper-BenchCode51.559.0+7.5
DTBenchGeneral QA89.997.3+7.5
GBENCH One-shot Coding GScoreGames48.055.4+7.4
IOI (Vals v2)Code40.347.6+7.3
Terminal-Bench 2.1 Extended (Mercor held-out)Code38.445.5+7.1
ARC-AGI-1 Semi-PrivateGeneral QA94.587.5−7.0
Legal Research BenchLegal41.348.1+6.7
DeepSWE v1.1Code74.267.5−6.7
Bug Hunt BenchCode20.627.3+6.7
SpeciEval (Refusal-Aware Retries, Historical)Safety97.490.8−6.5
SpeciEval (Pooled English Runs)Safety97.490.8−6.5
SimpleQA Verified (Epoch anti-abstention)General QA42.949.3+6.4
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science5.211.5+6.4
AA-AnalystAgent (pass^5)Professional Work35.041.3+6.3
CyberBench v1.1 OverallCybersecurity73.767.7−6.0
Arena Alignment IndexSafety74.880.8+6.0
GDPval-AA v2Professional Work16321753+5.9
ARC-AGI-2General QA72.967.1−5.8
Furniture Assembly BenchmarkMultimodal34.240.0+5.8
APEX-SWECode51.957.7+5.8
Conceptual Reasoning IndexIndexes58.462.5+5.6
AA-LCR Extended (Mercor held-out)Search70.865.2−5.6
Terminal-Bench 4.0 (Artificial Analysis)Code26.821.2−5.6
Excel Modeling BenchmarkFinance & Accounting57.262.7+5.5
SpatialBench V0 VerifiedScience75.469.9−5.5
SWE-bench MultilingualCode98.292.7−5.5
CritPt (avg@5)Science14.319.7+5.4
BuseyBench SVGMultimodal4.14.6+5.2
LMArena Agent Arena Code Bash ReliabilityAgents-0.210.1+5.1
LMArena Agent Arena Work Net Improvement (Steering Burden)Agents2.50.4−5.1
Harvey HLAB — Criteria Pass RateLegal87.592.5+5.0
FrontierCode 1.1 Extended - Pass RateCode62.067.0+5.0
FrontierMath Tier 4 (v2) - PrivateMath26.831.7+4.9
Artificial Analysis Intelligence Index v4.3Indexes39.544.4+4.9
Artificial Analysis Intelligence Index v4.3.2Indexes39.544.3+4.9
LMArena Agent Arena Work Bash ReliabilityAgents0.19.7+4.8
RareDiseaseBench V1 (35 evaluations)Science0.04.8+4.8
SWE-bench Verified Extended (Mercor held-out)Code82.778.0−4.7
Wiz Cyber Model Arena v3.9Cybersecurity50.245.5−4.7
FrontierCode 1.1 Extended - ScoreCode56.661.3+4.7
Design Arena Overall Code CategoriesCode13161288−4.6
SciCode 1.0.1 (Artificial Analysis)Code51.956.5+4.6
BrowseComp (Mercor 130-question subset)Search79.484.0+4.6
MacroscopeBenchCode70.575.1+4.6
ScBench V1 Clean RerunsScience54.759.1+4.5
BioMysteryBench (Vals Terminus 2)Science67.872.2+4.4
HealthBench ProfessionalHealth & Medicine52.948.5−4.4
LMArena Agent Arena Chat Bash ReliabilityAgents-1.07.2+4.1
SpatialBench V1 Clean RerunsScience63.267.3+4.1
LMArena Agent Arena Work Confirmed SuccessAgents5.7-2.4−4.1
LMArena Agent Arena Chat Confirmed SuccessAgents4.9-3.2−4.0
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine14.718.7+4.0
CadQueryEval (Corrected Geometry)Code76.072.0−4.0
GeneBench-Pro Public Release SubsetScience11.015.0+4.0
Terminal-Bench 3 - release-reported accuracyCode30.026.0−4.0
BioSecBench-Surveillance V1 Clean RerunsScience51.547.5−3.9
AA-PersonalityBench Conviction indexChat & Writing96.292.4−3.8
Terminal-Bench 2.1Code74.578.3+3.7
Senior SWE-BenchCode22.726.3+3.6
MedCodeHealth & Medicine41.244.7+3.5
ACCoRDGeneral QA68.171.6+3.5
SciCode (main problems)Science11.214.6+3.4
LMArena Agent Arena Chat Net Improvement (Steering Burden)Agents1.60.3−3.3
BioSecBench-Function V0Science38.742.0+3.3
LMArena Agent Arena SteerabilityAgents-3.33.3+3.3
LMArena Text Arena EloChat & Writing14751456−3.2
SWE Atlas - Codebase QnACode53.550.3−3.2
BioSecBench-Function V1 Clean RerunsScience40.844.0+3.2
YC-BenchGames16572361568241−3.1
LLM Debate BenchmarkChat & Writing15821604+3.1
LiveBench 2.0Agents78.074.9−3.1
AdvancedIFChat & Writing84.080.9−3.1
LMArena Agent Arena Code Confirmed SuccessAgents6.60.4−3.1
LMArena Agent Arena Code Net Improvement (Steering Burden)Agents2.71.4−3.1
LegalBenchLegal83.386.3+3.0
ScBench-Long V0 VerifiedScience27.324.2−3.0
ProofBench v1.1Math54.051.0−3.0
Humanity's Last ExamGeneral QA36.839.7+2.9
Terminal-Bench-Science 0.1Science9.06.2−2.9
CharXiv (descriptive and reasoning combined)Multimodal90.192.9+2.8
ReactBench (39 tasks): Writing ReactCode47.044.3−2.6
LMArena Agent Arena Work Praise vs ComplaintAgents3.7-1.5−2.6
GDP.pdfProfessional Work19.817.2−2.6
Public Benefits Bench v1.1Professional Work64.366.8+2.6
GDPval-AA v2.1Professional Work16001658+2.6
Terminal-Bench 4.0Code19.717.2−2.5
Humanity's Last Exam DiamondGeneral QA19.121.5+2.4
APEX-1 (Mercor, mean score)Professional Work65.567.7+2.2
LMArena Agent Arena Chat Praise vs ComplaintAgents3.1-1.1−2.1
LMArena Agent Arena Code Praise vs ComplaintAgents3.0-1.2−2.1
CWE-bench v1 Programmatic Pass@1Cybersecurity55.057.0+2.0
AA-PersonalityBench Expressiveness indexChat & Writing61.963.8+1.9
AA-PersonalityBench Slop indexChat & Writing37.235.3−1.9
AutomationBench-AA 1.0.6Agents68.967.0−1.9
EpiBench V1 Clean RerunsScience25.523.6−1.9
ScBench-Long V1 Clean RerunsScience14.816.5+1.7
AIME 2025Math98.3100.0+1.7
SWE-bench VerifiedCode82.984.6+1.7
Access to Justice BenchmarkLegal72.971.3−1.5
LMCAGeneral QA47.048.5+1.5
NeuroscienceBench V1 (91 evaluations)Science22.020.5−1.5
FrontierMath Tiers 1-3 (v2) - PrivateMath67.466.0−1.4
Vals Index v2Indexes57.959.2+1.3
MMLU ProGeneral QA88.289.4+1.2
Epoch Capabilities Index (ECI)Indexes156158+1.2
ARC-AGI-3 (Semi-Private)General QA3.32.1−1.2
LMArena Agent Arena Bash RecoveryAgents7.95.7−1.1
Code MigrationCode45.644.6−1.1
LMArena Agent Arena Chat Steering BurdenAgents-0.51.6+1.0
MedScribeHealth & Medicine85.586.5+1.0
MMMU-Pro Extended (Mercor held-out)Multimodal25.026.0+1.0
OTIS Mock AIME 2024-2025Math98.399.2+0.8
Arena Image-to-WebDevCode16101607−0.7
GDPval (Mercor 213-task rubric score)Professional Work86.786.0−0.7
MetagenomicsBench V0Science38.038.7+0.7
MetagenomicsBench V1Science38.038.7+0.7
MineBench — Bradley-Terry RatingGames18221811−0.7
Design Creativity BenchCode43.643.4−0.6
LMArena Agent Arena Work Steering BurdenAgents1.50.4−0.6
HLE Extended (Mercor held-out)General QA45.345.8+0.5
ProgramBenchCode0.51.0+0.5
LMArena Agent Arena Code Steering BurdenAgents2.31.3−0.5
Artificial Analysis Progress Index v1Indexes320322+0.5
Bedside BenchHealth & Medicine82.983.3+0.4
Immigration BenchLegal65.666.0+0.3
LMArena Code Arena WebDevCode16211618−0.3
GameBench 2Games68.468.7+0.3
ALE-BenchCode18291822−0.2
Finance Agent v2Finance & Accounting53.553.7+0.2
TCSExamMath40.440.5+0.1
Vision Arena OverallMultimodal12631263+0.1
LMArena Agent Arena Code Tool HallucinationAgents0.50.60.0
LMArena Agent Arena Work Tool HallucinationAgents0.50.60.0
LMArena Agent Arena Chat Tool HallucinationAgents0.30.30.0
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 - Corporate M&ALegal20.020.00.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal20.020.00.0
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 - Healthcare Life SciencesLegal20.020.00.0
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 - White Collar Defense InvestigationsLegal0.00.00.0
MedXpertQA MM Extended (Mercor held-out)Health & Medicine48.448.40.0
SpatialBench-Long V1 Clean RerunsScience26.226.20.0
Terminal-Bench Science 0.1 (Vals)Science4.34.30.0
TissueMicroscopyBench V1 (49 evaluations)Science2.72.70.0

DeepSeek V4.1 FlashvsGrok 4.6normalised score marginsData Nov 15, 2026Data: Vals AI, benchmarks.bio, Arena and 42 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…