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 Knowledge. Every model the catalog carries a published Decision Index Knowledge value for, ranked by that value.
Decision Index 0.4 publishes chance-corrected knowledge scores for 209 evaluated systems.
120 models with a published Decision Index Knowledge 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 | 51.4 |
| 2 | CHQuyet-1.0-Large | chinhnc | 47.0 |
| 3 | TAJev 1.14 | TypeSafe AI | 46.6 |
What the benchmark measures and how its value is published, from the source’s own record.
Decision Index Knowledge (Decision Index 0.4 Knowledge) is an external benchmark published by Decision Index leaderboard. Decision Index 0.4 publishes chance-corrected knowledge scores for 209 evaluated systems. Its evaluation set is described by the source as knowledge area of the Decision Index evaluation. The source rates its difficulty as knowledge 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 |
| NLJev |
| Not listed |
| 46.2 |
| 5 | Perplexity | 46.0 |
| 6 | BAelephant-1 | Baseten | 45.8 |
| 6 | Perplexity | 45.8 |
| 8 | MADecider chat · Gemma-4-31B | Mapika | 44.8 |
| 8 | NLdeck31b (Gemma 4 31B) | Not listed | 44.8 |
| 10 | NLFastino GLiDE no-thinking (28B) | Not listed | 42.5 |
| 11 | SUSurogate Rune 26B-A4B v3 | Surogate | 42.0 |
| 12 | JUXor 26B-A4B | juspay | 41.8 |
| 13 | PIBlink v0.3 26B-A4B (NVFP4) | PixilabAI | 40.7 |
| 14 | FIJebadiah 27B | frontier-infra | 39.3 |
| 15 | AUJEV-27B | autotrust | 39.0 |
| 16 | NLTorchcast Decision 27B | Not listed | 38.9 |
| 17 | AUGEV-26B-Decide | autotrust | 38.7 |
| 18 | JAKev 27B | jaredpalmer | 38.3 |
| 19 | NLDecider chat · Qwen3.6-27B | Not listed | 38.0 |
| 20 | NLvLLM-SR Decision 2.0 Vega 27B | Not listed | 37.8 |
| 21 | CAEikos-27B | caiovicentino1 | 37.6 |
| 22 | NLJADE | Not listed | 36.9 |
| 23 | CLClef | Cloudflare | 36.3 |
| 24 | NADrex v2.0 | Nace.AI | 36.0 |
| 25 | KIJPT-35B-A3B | kirp | 35.8 |
| 25 | NLXor 1.2 | Not listed | 35.8 |
| 27 | ANAJev lora5 (Gemma 4 12B) | andyzhang232 | 35.5 |
| 28 | CLClef-Omni | Cloudflare | 34.1 |
| 29 | FAsimple-jev · Qwen3.8-27B | Featherless AI | 33.9 |
| 30 | KSreflex-27b | kshetrajna12 | 33.5 |
| 31 | BEBespoke Nimble 9B v3 | bespokelabs | 32.9 |
| 32 | NLJoshuaSP diffusiongemma (open-jev) | Not listed | 32.7 |
| 33 | CLClef-Flash | Cloudflare | 32.5 |
| 34 | ALJevK5 9B v0.3.3 | alibiserikbay | 32.2 |
| 35 | MAdecider-35b-a3b | Mapika | 32.0 |
| 35 | Perplexity | 32.0 |
| 37 | NLdecisio 0.4.0 (Qwen3.6-35B-A3B) | Not listed | 31.2 |
| 37 | NADrex v1.5 | Nace.AI | 31.2 |
| 37 | ERWinnow-12B Q8 | Eldan Ring | 31.2 |
| 40 | NLvLLM-SR Decision 2.0 Lux 9B | Not listed | 30.3 |
| 41 | NLdecisio 0.4.0 (Gemma 4 12B) | Not listed | 30.0 |
| 42 | AUJEV-9B | autotrust | 29.8 |
| 43 | KIJPT-9B | kirp | 29.1 |
| 44 | ELjiwo 4B | eljiwo | 28.9 |
| 45 | NLKev 9B v2 | Not listed | 28.6 |
| 46 | EVezjev 4B s2 | everettjf | 28.5 |
| 47 | NLvLLM-SR Decision 2.0 Nox 4B | Not listed | 28.4 |
| 48 | NLHopper 12B (trained) | Not listed | 27.8 |
| 49 | KIJPT-4B | kirp | 27.2 |
| 50 | BEBespoke Nimble 9B v2 | bespokelabs | 26.6 |
| 51 | HOHopper (G) 1.2 | HopitAI | 26.5 |
| 52 | MADecider 4B | Mapika | 26.4 |
| 53 | KPKas 4B | kpiya | 26.2 |
| 54 | MIJet v6.2 | michaljach | 25.9 |
| 55 | NLJevK5 v0.3 | Not listed | 25.8 |
| 56 | SASieve 9B | sthanika-ai | 25.2 |
| 57 | NLJevfire | Not listed | 24.5 |
| 58 | NLopen-jev (pngwn) | Not listed | 23.9 |
| 59 | AKJev-Omni | akhilaaa3 | 23.4 |
| 60 | JAKev 4B r10 | jaredpalmer | 23.3 |
| 60 | TONeoHorse-Jev-4B | TokenRhythm | 23.3 |
| 62 | NLmmastrac diffusiongemma (vLLM PR 57250) | Not listed | 23.2 |
| 63 | internlm | 22.9 |
| 63 | NLrazorback16 openjev diffusiongemma (NVFP4, vLLM) | Not listed | 22.9 |
| 65 | TOTev1-4B-experimental | togethercomputer | 22.8 |
| 66 | AMJevstral 8B | AmirBraham | 22.5 |
| 67 | MAdjev | Maisa | 22.4 |
| 68 | SHRSI-Jev v5.0 VL 3B | shgao | 21.8 |
| 69 | W(metask-jev-4b | Wayfind (metask-ai) | 21.4 |
| 69 | NLSolomon v1.1 | Not listed | 21.4 |
| 71 | IAlev | interfaze-ai | 20.8 |
| 72 | NLIntelif Qwen3-4B | Not listed | 20.3 |
| 73 | NLLev1 3.7B | Not listed | 20.1 |
| 73 | ELWinnow-E4B | EldanRing | 20.1 |
| 75 | NLopenvons | Not listed | 19.9 |
| 76 | NLRyotide Qwen v0.3.1 | Not listed | 18.2 |
| 77 | NLmini-jev | Not listed | 18.1 |
| 78 | NLJev-Style 2B v3 | Not listed | 17.9 |
| 79 | DAGevva E4B | davidburhans | 16.7 |
| 80 | NLJobe | Not listed | 16.2 |
| 81 | MAdecider-2b | Mapika | 15.4 |
| 82 | NLSemIf | Not listed | 15.3 |
| 83 | NLEXAONE-4.0-1.2B-JEV v0.3 | Not listed | 15.2 |
| 84 | NLthis-that 1.2 | Not listed | 14.3 |
| 85 | DAGevva E2B | davidburhans | 13.9 |
| 86 | NLvLLM-SR Decision 2.0 Sol 2B | Not listed | 13.0 |
| 87 | SASieve 2B | sthanika-ai | 12.4 |
| 88 | ELjiwo 0.8B | eljiwo | 11.3 |
| 89 | LUDinah-0 | Lukitaduarte | 9.7 |
| 90 | internlm | 9.5 |
| 91 | CAKodiak v0.2 1B (accuracy) | cortex-agent-llc | 9.4 |
| 92 | OMOneJev 0.8B | OmniJev | 9.0 |
| 93 | NLSifr 0.8B v3.1 | Not listed | 8.0 |
| 94 | HOBekko System One v0 400M | hotchpotch | 7.9 |
| 95 | NLvLLM-SR Decision 2.0 Eos 0.8B | Not listed | 7.6 |
| 96 | NLKev 0.8B r15 | Not listed | 7.4 |
| 97 | KIJPT-0.8B | kirp | 6.6 |
| 97 | MOLFM2.5-2.6B-RLCD | monotykamary | 6.6 |
| 99 | internlm | 6.5 |
| 100 | NLVon 1.3 | Not listed | 6.2 |
| 101 | NLvLLM-SR Decision 2.0 Kai 0.6B | Not listed | 6.0 |
| 102 | ALJevK5 Lite | alibiserikbay | 4.7 |
| 103 | TOTev1-0.8B-experimental | togethercomputer | 4.4 |
| 104 | SUJulia 1 | SupersonicLabs | 4.1 |
| 105 | HLBosun v3.1 1.7B | Hanno-Labs | 3.0 |
| 105 | NLJev-Style 0.8B v3 | Not listed | 3.0 |
| 107 | MOLavoir | moganai | 2.6 |
| 108 | NLsystem-one-gemma | Not listed | 2.3 |
| 109 | NLVerdict | Not listed | 2.1 |
| 110 | FAGLiNER2.5-Decide | fastino | 2.0 |
| 111 | HAQwen-2.5-1B-RLCD | harshatheg | 1.6 |
| 112 | CILaya | Convai Innovations | 1.4 |
| 113 | HOBekko System One v0 68M | hotchpotch | 0.9 |
| 113 | LMjeff | Logan Markewich | 0.9 |
| 115 | CLBosun v3.1 0.6B | Clause Logic | 0.4 |
| 115 | NLDM-JEPA 1.1 | Not listed | 0.4 |
| 117 | CLCLM-v0.1-8B | Contrastive-LM | 0.1 |
| 118 | HOBekko System One v0 17M | hotchpotch | 0.0 |
| 118 | NOLFM2.5-350M-RLCD | notnotsamuel | 0.0 |
| 118 | NLZev | 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.