General
Artificial Analysis Decision Index leaderboard
As of 2026-11-18, the highest measured score on Artificial Analysis Decision Index is 74.8% by d1.2-ultra. 89 more models have estimated scores, calibrated from the benchmarks they were measured on.
Measured scores: benchlm.ai.
| # | Model | Score | Source |
|---|---|---|---|
| 1 | d1.2-ultra | 74.8% | measured |
| 2 | Drex Max v1.1 | 72.6% | measured |
| 3 | GLiDE v1.2 | 72.5% | measured |
| 4 | mammoth-1 | 71.4% | measured |
| 5 | d1-ultra | 70.6% | measured |
| 6 | Perplexity Decider v1.3 27B | 70.4% | measured |
| 7 | Perplexity Decider v1.2 27B | 69.8% | measured |
| 8 | Drex Max | 69.1% | estimated ± 1.3 pp, high confidence |
| 9 | d1.2 | 68.7% | measured |
| 10 | SubQ Snap 1.1 | 67.9% | measured |
| 11 | deck31b (Gemma 4 31B) | 67.3% | estimated ± 1.3 pp, high confidence |
| 12 | Quyet-1.0-Large | 67.1% | estimated ± 1.3 pp, high confidence |
| 13 | Decider chat · Gemma-4-31B | 66.8% | estimated ± 1.3 pp, high confidence |
| 14 | elephant-1 | 66.4% | measured |
| 15 | Jev 1.14 | 66.2% | measured |
| 16 | Gemini Decision 1 | 66.0% | measured |
| 17 | Torchcast Decision 27B | 65.2% | estimated ± 1.3 pp, high confidence |
| 18 | Kev 27B | 64.4% | estimated ± 1.3 pp, high confidence |
| 19 | Surogate Rune 26B-A4B v3 | 64.3% | estimated ± 1.3 pp, high confidence |
| 20 | Perplexity Decider Flash v1.3 4B | 64.1% | measured |
| 21 | Celeris-1 Decision | 64.0% | estimated ± 2.9 pp, high confidence |
| 22 | GLiDE | 64.0% | estimated ± 2.9 pp, high confidence |
| 23 | Eikos-27B | 63.9% | estimated ± 1.3 pp, high confidence |
| 24 | JEV-27B | 63.8% | estimated ± 1.3 pp, high confidence |
| 25 | Blink v0.3 26B-A4B (NVFP4) | 63.8% | estimated ± 1.3 pp, high confidence |
| 26 | vLLM-SR Decision 2.0 Vega 27B | 63.7% | estimated ± 1.3 pp, high confidence |
| 27 | JADE | 62.6% | estimated ± 1.3 pp, high confidence |
| 28 | Decider chat · Qwen3.6-27B | 62.5% | estimated ± 1.3 pp, high confidence |
| 29 | GEV-26B-Decide | 62.3% | estimated ± 1.3 pp, high confidence |
| 30 | simple-jev · Qwen3.8-27B | 62.2% | estimated ± 1.3 pp, high confidence |
| 31 | Mercury Decide | 62.1% | measured |
| 32 | Microsoft-Decision-1 | 62.1% | measured |
| 33 | Celeris-1.1 Decision | 62.0% | measured |
| 34 | JPT-35B-A3B | 61.7% | estimated ± 1.3 pp, high confidence |
| 35 | Drex v2.0 | 61.6% | measured |
| 36 | Xor 1.2 | 61.5% | estimated ± 1.3 pp, high confidence |
| 37 | d1 | 60.8% | measured |
| 38 | Clef | 60.3% | measured |
| 39 | GPT-6 Luna (Decisions) | 60.2% | measured |
| 40 | Perplexity Decider Flash v1.2 4B | 60.1% | measured |
| 41 | Bespoke Nimble 9B v3 | 59.6% | estimated ± 1.3 pp, high confidence |
| 42 | decisio 0.4.0 (Qwen3.6-35B-A3B) | 58.9% | estimated ± 1.3 pp, medium confidence |
| 43 | JoshuaSP diffusiongemma (open-jev) | 58.8% | estimated ± 1.3 pp, high confidence |
| 44 | Clef-Omni | 57.8% | measured |
| 45 | decisio 0.4.0 (Gemma 4 12B) | 57.6% | estimated ± 1.3 pp, medium confidence |
| 46 | decider-35b-a3b | 57.4% | estimated ± 1.3 pp, high confidence |
| 47 | GLiNER2.5-Decide | 57.4% | estimated ± 2.6 pp, medium confidence |
| 48 | Bosun v3.1 1.7B | 57.4% | estimated ± 2.6 pp, medium confidence |
| 49 | Laya | 57.4% | estimated ± 2.6 pp, medium confidence |
| 50 | Winnow-12B Q8 | 57.1% | estimated ± 1.3 pp, medium confidence |
| 51 | Drex v1.5 | 56.7% | estimated ± 1.3 pp, medium confidence |
| 52 | JevK5 9B v0.3.3 | 56.7% | estimated ± 1.3 pp, medium confidence |
| 53 | Clef-Flash | 56.6% | estimated ± 1.3 pp, medium confidence |
| 54 | vLLM-SR Decision 2.0 Lux 9B | 56.5% | estimated ± 1.3 pp, medium confidence |
| 55 | Hopper 12B (trained) | 55.5% | estimated ± 1.3 pp, medium confidence |
| 56 | JPT-9B | 55.2% | estimated ± 1.3 pp, medium confidence |
| 57 | ezjev 4B s2 | 54.9% | estimated ± 1.3 pp, medium confidence |
| 58 | JEV-9B | 54.5% | estimated ± 1.3 pp, medium confidence |
| 59 | Jevfire | 53.6% | estimated ± 1.3 pp, medium confidence |
| 60 | Kev 9B v2 | 53.5% | estimated ± 1.3 pp, medium confidence |
| 61 | vLLM-SR Decision 2.0 Nox 4B | 53.4% | estimated ± 1.3 pp, medium confidence |
| 62 | Bespoke Nimble 9B v2 | 53.3% | estimated ± 1.3 pp, medium confidence |
| 63 | Jev-Omni | 52.9% | estimated ± 1.3 pp, medium confidence |
| 64 | Sieve 9B | 52.6% | estimated ± 1.3 pp, medium confidence |
| 65 | Hopper (G) 1.2 | 52.3% | estimated ± 1.3 pp, medium confidence |
| 66 | JPT-4B | 52.3% | estimated ± 1.3 pp, medium confidence |
| 67 | Solomon v1.1 | 51.8% | estimated ± 1.3 pp, medium confidence |
| 68 | Decider 4B | 50.3% | estimated ± 1.3 pp, medium confidence |
| 69 | JevK5 v0.3 | 50.2% | estimated ± 1.3 pp, medium confidence |
| 70 | Intern-Decision-4B | 50.1% | estimated ± 1.3 pp, medium confidence |
| 71 | djev | 49.6% | estimated ± 1.3 pp, medium confidence |
| 72 | NeoHorse-Jev-4B | 49.5% | estimated ± 1.3 pp, medium confidence |
| 73 | razorback16 openjev diffusiongemma (NVFP4, vLLM) | 49.5% | estimated ± 1.3 pp, medium confidence |
| 74 | Kas 4B | 49.0% | estimated ± 1.3 pp, medium confidence |
| 75 | Kev 4B r10 | 48.2% | estimated ± 1.3 pp, medium confidence |
| 76 | mmastrac diffusiongemma (vLLM PR 57250) | 47.8% | estimated ± 1.3 pp, medium confidence |
| 77 | lev | 47.4% | estimated ± 1.3 pp, medium confidence |
| 78 | Tev1-4B-experimental | 47.3% | estimated ± 1.3 pp, medium confidence |
| 79 | Lev1 3.7B | 47.0% | estimated ± 1.3 pp, medium confidence |
| 80 | open-jev (pngwn) | 46.3% | estimated ± 1.3 pp, medium confidence |
| 81 | Jevstral 8B | 46.3% | estimated ± 1.3 pp, medium confidence |
| 82 | Bekko System One v0 68M | 46.0% | estimated ± 2.7 pp, medium confidence |
| 83 | metask-jev-4b | 45.8% | estimated ± 1.3 pp, medium confidence |
| 84 | Winnow-E4B | 45.4% | estimated ± 1.3 pp, medium confidence |
| 85 | Intelif Qwen3-4B | 44.7% | estimated ± 1.3 pp, medium confidence |
| 86 | Ryotide Qwen v0.3.1 | 44.1% | estimated ± 1.3 pp, medium confidence |
| 87 | Qwen-2.5-1B-RLCD | 43.9% | estimated ± 2.7 pp, medium confidence |
| 88 | openvons | 42.9% | estimated ± 1.3 pp, medium confidence |
| 89 | Jobe | 42.1% | estimated ± 1.3 pp, medium confidence |
| 90 | Gevva E4B | 41.6% | estimated ± 1.3 pp, medium confidence |
| 91 | SemIf | 39.1% | estimated ± 1.3 pp, medium confidence |
| 92 | mini-jev | 37.9% | estimated ± 1.3 pp, medium confidence |
| 93 | this-that 1.2 | 37.3% | estimated ± 1.3 pp, medium confidence |
| 94 | decider-2b | 37.0% | estimated ± 1.3 pp, medium confidence |
| 95 | Gevva E2B | 36.6% | estimated ± 1.3 pp, medium confidence |
| 96 | vLLM-SR Decision 2.0 Sol 2B | 35.6% | estimated ± 1.3 pp, medium confidence |
| 97 | EXAONE-4.0-1.2B-JEV v0.3 | 34.6% | estimated ± 1.3 pp, medium confidence |
| 98 | Sieve 2B | 33.9% | estimated ± 1.3 pp, medium confidence |
| 99 | Intern-Decision-2B | 31.2% | estimated ± 1.3 pp, medium confidence |
| 100 | Kodiak v0.2 1B (accuracy) | 28.1% | estimated ± 1.9 pp, medium confidence |
| 101 | vLLM-SR Decision 2.0 Eos 0.8B | 27.8% | estimated ± 1.3 pp, medium confidence |
| 102 | Bekko System One v0 400M | 27.6% | estimated ± 1.3 pp, medium confidence |
| 103 | Sifr 0.8B v3.1 | 24.9% | estimated ± 1.9 pp, medium confidence |
| 104 | Kev 0.8B r15 | 23.4% | estimated ± 1.9 pp, medium confidence |
| 105 | JPT-0.8B | 21.4% | estimated ± 1.9 pp, medium confidence |
| 106 | LFM2.5-2.6B-RLCD | 21.4% | estimated ± 1.9 pp, medium confidence |
| 107 | Intern-Decision-0.8B | 21.1% | estimated ± 1.9 pp, medium confidence |
| 108 | Von 1.3 | 20.3% | estimated ± 1.9 pp, medium confidence |
| 109 | vLLM-SR Decision 2.0 Kai 0.6B | 19.8% | estimated ± 1.9 pp, medium confidence |
| 110 | JevK5 Lite | 16.2% | estimated ± 1.9 pp, medium confidence |
| 111 | Tev1-0.8B-experimental | 15.3% | estimated ± 1.9 pp, medium confidence |
All leaderboards
Agentic · terminal
- AA Terminal-Bench 2.1
- AA Terminal-Bench 4.0
- Terminal-Bench 4.0
- Terminal-Bench 2.0
- Terminal-Bench 2.1
- Terminal-Bench 3.0
- Terminal-Bench Hard
- Terminal-Bench-Science 0.1
- Terminal-Bench 4.1
- Terminal-Bench 2.1 (Vals AI)
Agentic · tools
- AA Agentic Index
- AA-AnalystAgent
- AA AutomationBench
- CWE-Bench-AA
- AA Cyber Index
- DeepsecBench-AA
- AA EnterpriseOps-Gym
- GDP.pdf
- AA Harvey LAB v1.0
- Harvey LAB-AA v1.1
- AA ITBench
- AA Tau3 Banking
- Agents' Last Exam
- APEX-Agents
- APEX-Agents-AA
- ApprenticeBench
- AutomationBench
- BFCL v4
- BrowseComp
- Claw-Eval
- CWE-bench v1
- CWE-bench
- CyberGym
- DeepPlanning
- DeepSearchQA
- DRACO
- ExploitGym
- Finance Agent v2
- GDPval-AA
- Gert Labs
- Hermes Bench
- Hermes Index
- HLE w/ tools
- JobBench
- LifeVal
- MazeBench
- MCP Atlas
- MCP-Tasks
- MKA
- OSWorld 2.1
- OSWorld 2.1 Offline
- OSWorld-Verified
- OSWorld 2.0
- PinchBench
- QwenClawBench
- ResearchClawBench
- SkillsBench
- τ²-bench results
- τ³-bench results
- Toolathlon
- Toolathlon-Verified
- VITA-Bench
- WideResearch
Coding
- AA Coding Index
- AA LiveCodeBench
- AA-SciCode
- Bug Hunt Bench
- CursorBench 3.1
- CursorBench 3.2
- CursorBench 4.0
- DeepSWE
- FrontierCode 1.1 Main
- FrontierCode 2.0
- FrontierCode 1.1 Extended
- FrontierSWE v2
- LiveCodeBench
- LiveCodeBench Pro
- LiveCodeBench v6
- NL2Repo
- OpenHarmony Bench v1.2
- OpenHarmony Bench v1.0
- PostTrainBench
- PostTrainBench v1
- PostTrainBench v1.1
- ProgramBench
- React Native Evals
- ScarfBench
- SciCode
- SWE-Atlas Codebase QnA
- SWE-bench Pro
- SWE-bench Verified
- SWE-Rebench
- SWE-Marathon
- Vals LiveCodeBench
- Vals SWE-bench
- Vibe Code Bench
- VulcanBench Frontier v4
- VulcanBench v3
Math
- AIME 2024
- AIME 2025
- AIME26
- FrontierMath v2 (Tier 4)
- FrontierMath v2 (Tiers 1-3)
- HMMT Feb 2025
- HMMT Feb 2026
- HMMT Nov 2025
- IMOAnswerBench
- MATH-500
- MMAnswerBench
Knowledge & reasoning
- AA-GPQA Diamond
- HealthBench Professional (length-adjusted, Artificial Analysis)
- HealthBench Professional (raw, Artificial Analysis)
- AA-HLE
- AA MMLU-Pro
- ARC-AGI-2
- ARC-AGI-3 Provider Adapter
- ARC-AGI-1
- ARC-AGI-3
- Artificial Analysis Intelligence Index
- BBH
- BioBasicsBench
- BioMysteryBench (human-difficult)
- BioMysteryBench (human-solvable)
- C-Eval
- CritPt
- GPQA
- GPQA Diamond
- HealthBench (raw)
- HealthBench Hard
- HealthBench (length-adjusted)
- HealthBench Professional
- HealthBench Professional (raw)
- HLE
- HLE w/o tools
- HLE-Verified
- JevBench 1.4
- LABBench2
- MedXpertQA (Text)
- MMLU
- MMLU-Pro
- MMLU-Pro (Arcee)
- MMLU-Redux
- MMMLU
- AA-Omniscience Accuracy
- SimpleQA
- STEMBench
- SuperGPQA
- Vals GPQA Diamond
- Vals MMLU-Pro
Instruction following
Multilingual
Vision & documents
- AA-MMMU-Pro
- AI2D_TEST
- BabyVision
- BabyVision w/ Python
- BenchCAD Vision2Code (no tools)
- BenchCAD Vision2Code (tools)
- CC-OCR
- Chartography (no tools)
- Chartography (tools)
- CharXiv
- CharXiv w/o tools
- CountBench
- DynaMath
- ERQA
- LVBench
- MathVision
- MathVision w/ Python
- MedXpertQA (MM)
- MLVU (M-Avg)
- MMMU
- MMMU-Pro
- MMMU-Pro w/ Python
- MMVU
- OCRBench V2
- OfficeQA
- OfficeQA Pro
- OmniDocBench 1.5
- ParseBench-1
- ParseBench-2
- PerceptionBench
- RealWorldQA
- RefCOCO (avg)
- ScreenSpot Pro
- SimpleVQA
- SWE-bench Multimodal
- Video-MME (w/o subtitle)
- Video-MME (with subtitle)
- VideoMMMU
- V*
- We-Math
- ZeroBench
Long context
General
- HLE-MultiChoice (Decision Index)
- Artificial Analysis Decision Index v1.0
- Artificial Analysis Decision Index
- When2Call (Decision Index)
- WildGuardTest (Decision Index)
- BabaIsBench
- BinaryAudit
- BlenderWorld (text-only)
- Chess Puzzles
- CompileBench
- Cookie-Bench
- DAYJOB
- DAYJOB Healthcare
- Decision Index Full Score
- EBR-bench
- Epoch Automation Reports
- ExploitBench
- ExploitBench Internal Port
- FACTS Multimodal
- FailureSensorIQ
- FrontierPhysics
- Furniture Assembly
- Decision Index Arts
- Decision Index Knowledge
- Decision Index Language
- Decision Index Retrieval
- Decision Index Tools
- MaxxBench
- MirrorCode
- Mystery Game Puzzles
- NYT Connections
- OTelBench
- SEC-Bench Pro
- SkillsBench official leaderboard (BenchFlow)
- SpireClimb
- StealthMark