Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
external benchmark
Decision Index 0.4 Arts. Every model the catalog carries a published Decision Index Arts value for, ranked by that value.
Decision Index 0.4 publishes chance-corrected arts scores for 209 evaluated systems.
120 models with a published Decision Index Arts value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Chance-corrected area score |
|---|---|---|---|
| 1 | BAmammoth-1 | Baseten | 60.9 |
| 2 | BAelephant-1 | Baseten | 56.7 |
| 3 | Perplexity | 56.4 | |
What the benchmark measures and how its value is published, from the source’s own record.
Decision Index Arts (Decision Index 0.4 Arts) is an external benchmark published by Decision Index leaderboard. Decision Index 0.4 publishes chance-corrected arts scores for 209 evaluated systems. Its evaluation set is described by the source as arts area of the Decision Index evaluation. The source rates its difficulty as arts and human-taste decision tasks.
Each model's result is published by Decision Index leaderboard as chance-corrected area score. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 120 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.
| 4 |
| NADrex v2.0 |
| Nace.AI |
| 55.2 |
| 5 | NLFastino GLiDE no-thinking (28B) | Not listed | 53.1 |
| 6 | PIBlink v0.3 26B-A4B (NVFP4) | PixilabAI | 52.3 |
| 7 | NLTorchcast Decision 27B | Not listed | 52.2 |
| 8 | SUSurogate Rune 26B-A4B v3 | Surogate | 52.0 |
| 9 | Perplexity | 51.4 |
| 10 | NADrex v1.5 | Nace.AI | 50.1 |
| 11 | AUGEV-26B-Decide | autotrust | 48.6 |
| 12 | NLXor 1.2 | Not listed | 48.1 |
| 13 | TAJev 1.14 | TypeSafe AI | 47.7 |
| 14 | BEBespoke Nimble 9B v3 | bespokelabs | 47.2 |
| 15 | CAEikos-27B | caiovicentino1 | 47.1 |
| 16 | AUJEV-27B | autotrust | 46.9 |
| 16 | JUXor 26B-A4B | juspay | 46.9 |
| 18 | NLJev | Not listed | 46.2 |
| 19 | NLvLLM-SR Decision 2.0 Vega 27B | Not listed | 45.6 |
| 20 | FAsimple-jev · Qwen3.8-27B | Featherless AI | 45.3 |
| 21 | NLJADE | Not listed | 45.2 |
| 22 | JAKev 27B | jaredpalmer | 45.1 |
| 22 | CHQuyet-1.0-Large | chinhnc | 45.1 |
| 24 | Perplexity | 45.0 |
| 25 | KSreflex-27b | kshetrajna12 | 44.8 |
| 26 | CLClef | Cloudflare | 44.7 |
| 27 | MADecider chat · Gemma-4-31B | Mapika | 44.3 |
| 27 | NLdeck31b (Gemma 4 31B) | Not listed | 44.3 |
| 29 | FIJebadiah 27B | frontier-infra | 43.0 |
| 30 | NLDecider chat · Qwen3.6-27B | Not listed | 42.9 |
| 30 | MAdecider-35b-a3b | Mapika | 42.9 |
| 32 | ANAJev lora5 (Gemma 4 12B) | andyzhang232 | 42.3 |
| 33 | KIJPT-35B-A3B | kirp | 41.4 |
| 34 | NLJoshuaSP diffusiongemma (open-jev) | Not listed | 40.9 |
| 35 | NLdecisio 0.4.0 (Qwen3.6-35B-A3B) | Not listed | 40.2 |
| 36 | IAlev | interfaze-ai | 39.4 |
| 37 | CLClef-Flash | Cloudflare | 39.3 |
| 38 | NLdecisio 0.4.0 (Gemma 4 12B) | Not listed | 38.7 |
| 38 | NLLev1 3.7B | Not listed | 38.7 |
| 40 | CLClef-Omni | Cloudflare | 37.8 |
| 40 | EVezjev 4B s2 | everettjf | 37.8 |
| 42 | NLJevfire | Not listed | 37.6 |
| 42 | NLvLLM-SR Decision 2.0 Lux 9B | Not listed | 37.6 |
| 44 | MAdjev | Maisa | 37.4 |
| 45 | NLHopper 12B (trained) | Not listed | 37.3 |
| 46 | NLrazorback16 openjev diffusiongemma (NVFP4, vLLM) | Not listed | 35.9 |
| 47 | NLRyotide Qwen v0.3.1 | Not listed | 35.7 |
| 48 | ALJevK5 9B v0.3.3 | alibiserikbay | 35.1 |
| 48 | NLJobe | Not listed | 35.1 |
| 48 | ERWinnow-12B Q8 | Eldan Ring | 35.1 |
| 51 | AUJEV-9B | autotrust | 34.6 |
| 52 | MADecider 4B | Mapika | 34.5 |
| 53 | BEBespoke Nimble 9B v2 | bespokelabs | 34.2 |
| 54 | HOHopper (G) 1.2 | HopitAI | 34.0 |
| 55 | NLJevK5 v0.3 | Not listed | 33.5 |
| 56 | ELjiwo 4B | eljiwo | 33.2 |
| 56 | KIJPT-9B | kirp | 33.2 |
| 56 | NLmmastrac diffusiongemma (vLLM PR 57250) | Not listed | 33.2 |
| 59 | NLSemIf | Not listed | 33.0 |
| 60 | internlm | 32.9 |
| 61 | NLopenvons | Not listed | 32.7 |
| 62 | AKJev-Omni | akhilaaa3 | 31.9 |
| 62 | NLmini-jev | Not listed | 31.9 |
| 62 | NLvLLM-SR Decision 2.0 Nox 4B | Not listed | 31.9 |
| 65 | KPKas 4B | kpiya | 31.6 |
| 66 | MIJet v6.2 | michaljach | 31.1 |
| 67 | NLSolomon v1.1 | Not listed | 30.7 |
| 68 | KIJPT-4B | kirp | 30.6 |
| 69 | SASieve 9B | sthanika-ai | 29.7 |
| 70 | DAGevva E4B | davidburhans | 28.3 |
| 71 | ELWinnow-E4B | EldanRing | 28.2 |
| 72 | TOTev1-4B-experimental | togethercomputer | 27.3 |
| 73 | NLIntelif Qwen3-4B | Not listed | 27.2 |
| 74 | HOBekko System One v0 400M | hotchpotch | 24.6 |
| 75 | MAdecider-2b | Mapika | 24.4 |
| 76 | SASieve 2B | sthanika-ai | 23.6 |
| 77 | NLJev-Style 2B v3 | Not listed | 23.2 |
| 78 | JAKev 4B r10 | jaredpalmer | 23.0 |
| 79 | W(metask-jev-4b | Wayfind (metask-ai) | 22.2 |
| 80 | SHRSI-Jev v5.0 VL 3B | shgao | 22.1 |
| 81 | internlm | 21.4 |
| 82 | NLKev 9B v2 | Not listed | 21.3 |
| 83 | NLopen-jev (pngwn) | Not listed | 20.3 |
| 84 | NLthis-that 1.2 | Not listed | 19.1 |
| 85 | AMJevstral 8B | AmirBraham | 17.6 |
| 86 | CAKodiak v0.2 1B (accuracy) | cortex-agent-llc | 17.3 |
| 87 | HOBekko System One v0 68M | hotchpotch | 16.9 |
| 88 | NLvLLM-SR Decision 2.0 Sol 2B | Not listed | 16.3 |
| 89 | NLEXAONE-4.0-1.2B-JEV v0.3 | Not listed | 16.2 |
| 90 | TONeoHorse-Jev-4B | TokenRhythm | 15.5 |
| 91 | ELjiwo 0.8B | eljiwo | 13.5 |
| 92 | HAQwen-2.5-1B-RLCD | harshatheg | 13.2 |
| 93 | DAGevva E2B | davidburhans | 13.0 |
| 94 | MOLFM2.5-2.6B-RLCD | monotykamary | 11.4 |
| 95 | OMOneJev 0.8B | OmniJev | 10.3 |
| 96 | NLvLLM-SR Decision 2.0 Eos 0.8B | Not listed | 9.6 |
| 97 | NLSifr 0.8B v3.1 | Not listed | 8.7 |
| 98 | HLBosun v3.1 1.7B | Hanno-Labs | 8.5 |
| 99 | internlm | 8.2 |
| 99 | NLvLLM-SR Decision 2.0 Kai 0.6B | Not listed | 8.2 |
| 101 | NLVon 1.3 | Not listed | 8.0 |
| 102 | NLJev-Style 0.8B v3 | Not listed | 7.9 |
| 103 | KIJPT-0.8B | kirp | 7.8 |
| 104 | NLsystem-one-gemma | Not listed | 7.3 |
| 105 | LUDinah-0 | Lukitaduarte | 7.0 |
| 105 | CILaya | Convai Innovations | 7.0 |
| 107 | CLBosun v3.1 0.6B | Clause Logic | 6.6 |
| 108 | ALJevK5 Lite | alibiserikbay | 5.8 |
| 109 | TOTev1-0.8B-experimental | togethercomputer | 5.7 |
| 110 | FAGLiNER2.5-Decide | fastino | 5.1 |
| 111 | MOLavoir | moganai | 4.8 |
| 112 | NLKev 0.8B r15 | Not listed | 4.7 |
| 113 | NOLFM2.5-350M-RLCD | notnotsamuel | 4.3 |
| 114 | HOBekko System One v0 17M | hotchpotch | 3.9 |
| 115 | NLDM-JEPA 1.1 | Not listed | 3.6 |
| 116 | SUJulia 1 | SupersonicLabs | 2.4 |
| 117 | CLCLM-v0.1-8B | Contrastive-LM | 1.5 |
| 118 | NLZev | Not listed | 1.1 |
| 119 | LMjeff | Logan Markewich | 1.0 |
| 120 | NLVerdict | Not listed | 0.0 |
Evidence key: ObservedLast good
Rows are ordered by the value Decision Index leaderboard published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.