whatmodel.sh / decision / decider-2b (DecisionBench fa996cea)

decider-2b (DecisionBench fa996cea)

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

decider-2b (DecisionBench fa996cea) is an open-weights decision model from Mapika, released 19 September 2026. Its published size is 1.881825088B parameters. Licence: apache-2.0.

Catalogue identity notes

DecisionBench v1 comparison: Mapika/decider-2b, revision fa996cea58e1c1d8d1ab4d7124154f303b017f95; decision-model; adapter jev-compatible-server-v1; probability source option-slot-logit-softmax. Kept separate from newer or unpinned Decision Index recipes; standard prompt, not compact. No unverified runtime or OpenRouter price imported. Current repository metadata checked separately; an unpinned alias is not evidence that today’s weights reproduce this historical run.

Model information sources

Release date ↗ · Weights licence ↗ · Parameter count ↗

Published served parameter count; not a RAM or download-size measurement.

53.0±1.6#182 of 355 decision modelsfrom 29 benchmark results across 2 sources

$ how it compares

The decision models ranked either side of decider-2b (DecisionBench fa996cea), out of 357 with a score, then more from Mapika.

#modelscorewhereRAMreleased
182PurpleMIST-Flash-1.0 Olaverse Lab53±1model files
183JevAny-Qwen-4B (direct token) (JevBench mini) SimpleJev53±2source code
184decider-2b (DecisionBench fa996cea) Mapika53±2model files2026-09-19
185cua-s1-4b-0.2 (DecisionBench 16818868) cua-ai53±2model files
186Bespoke-Nimble-9B (JevBench mini) Bespoke Labs53±2source code
more from Mapika
43Decider 31B (NVFP4) Mapika66±1model files2026-10
85Decider chat · Gemma-4-31B (decision readout) Mapika62±1source code
110decider-35b-a3b (DecisionBench d9783c58) Mapika60±2model files2026-09-20

see all decision models · find the ones that fit your machine

$ what it's good at

workflows▁▁▁▂▁▂▂▆█▄▁▁ #61 of 114 · 99% of its score · 28 tests

Beats Clef-omni, Clef-flash, Bespoke Nimble 9B v3 here, though they rank higher overall. Trails InternLM Intern-Decision Qwen3.5-4B, StartLux-Decision v1.1 0.8B, JevK5 v0.3 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · action selection (all-request accuracy) submitted49.6% ✎▂▁▂▃▄▃▁▁▃█▇▄ beats 26/56medium · avg 51%▇▇▇▇▇▇▇▇ 19%at its level
DecisionBench v1 · routing (all-request accuracy) submitted73.0% ✎▁▁▁▁▂▂▁▁▁▁▃█ beats 37/56medium · avg 64%▇▇▇▇▇ 13%sharp · at its level
DecisionBench v1 · navigation (all-request accuracy) submitted65.0% ✎▁▁·▃▁▁▁▁▁▁▃█ beats 27/56medium · avg 58%▇▇▇▇ 11%sharp · at its level
DecisionBench v1 · record classification (all-request accuracy) submitted38.0% ✎▃▂▁▂▂▂▄▁▃▆██ beats 26/56medium · avg 39%▇▇▇ 8.4%at its level
DecisionBench v1 · verification (all-request accuracy) submitted57.0% ✎▁▁▅▁▃▂▂▁▁▇▇█ beats 27/56medium · avg 56%▇▇ 6.0%blunt · at its level
DecisionBench v1 · document record classification (all-request accuracy) submitted45.4% ✎▂▂·▁▂▄▃▆▄█▁▁ beats 30/56medium · avg 46%▇▇ 4.8%at its level
DecisionBench v1 · triage (all-request accuracy) submitted77.0% ✎▁·▁·▂▃▄▃▃█▄▂ beats 25/56easy · avg 78%▇▇ 4.4%sharp
DecisionBench v1 · completion assessment (all-request accuracy) submitted92.0% ✎▁▁▂▃▄·▂▁▁▆█▂ beats 40/56very easy · avg 86%▇▇ 4.4%sharp
20 more tests
testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · rubric scoring prioritization (all-request accuracy) submitted45.1% ✎▁▁▅▅▃▃▅█▃▃·▁ beats 31/56medium · avg 45%▇▇ 4.0%at its level
DecisionBench v1 · bounded extraction (all-request accuracy) submitted93.6% ✎▃▃▂·▁▁▁▁▁▁█▅ beats 34/56very easy · avg 86%▇▇ 3.9%sharp · noisy
Chao Ma · clean support routing96.0% ✎█····█·····█ beats 1/2very easy · avg 96%▇ 3.5%sharp · maxed out · small run
DecisionBench v1 · content classification (all-request accuracy) submitted78.0% ✎▁····▁▁▁▁▁▁█ beats 31/56easy · avg 73%▇ 3.3%typical
DecisionBench v1 · review scoring (all-request accuracy) submitted51.0% ✎▁······▂▃▅▇█ beats 29/56medium · avg 52%▇ 2.7%at its level
DecisionBench v1 · retrieval ranking (all-request accuracy) submitted61.0% ✎▁▁·▁▁▂▃▂▃▁▁█ beats 55/56medium · avg 35%▇ 2.1%at its level · noisy
DecisionBench v1 · target selection (all-request accuracy) submitted45.0% ✎▁▁·▁▁▁▃▁▂▁▁█ beats 42/56hard · avg 29%▇ 1.9%at its level · noisy
DecisionBench v1 · risk scoring (all-request accuracy) submitted65.0% ✎▁▁·▁▂▂▁▃▃█▇▃ beats 38/56medium · avg 62%▇ 1.8%blunt · at its level
DecisionBench v1 · reasoning (all-request accuracy) submitted42.6% ✎▂·▁▁▆█▅▃▃▂▁▁ beats 25/56medium · avg 52%▇ 1.1%at its level
DecisionBench v1 · retrieval verification (all-request accuracy) submitted85.7% ✎█▁▂▁▁▁▁▁▃▅▁▁ beats 38/56easy · avg 74%▇ 0.7%noisy
DecisionBench v1 · document workflows (all-request accuracy) submitted76.0% ✎█▄▃▁▁▃▂▂▁·▁▁ beats 44/56easy · avg 66%▇ 0.7%blunt · at its level
DecisionBench v1 · guardrails moderation (all-request accuracy) submitted63.2% ✎▁▁▁▂▆▃▃█···▁ beats 49/56medium · avg 52%▇ 0.5%blunt · at its level · noisy
DecisionBench v1 · safety gating (all-request accuracy) submitted86.0% ✎▂·····▆█▂▃▄▇ beats 34/56easy · avg 77%▇ 0.4%noisy
DecisionBench v1 · document classification (all-request accuracy) submitted21.0% ✎▂▁·▂▁▁▁▄▁▃▂█ beats 30/56hard · avg 21%▇ 0.3%blunt · noisy
DecisionBench v1 · semantic filtering (all-request accuracy) submitted81.0% ✎▁·▁▁·▁▁█▂▃█▂ beats 34/56easy · avg 71%▇ 0.3%blunt · at its level · noisy
DecisionBench v1 · content moderation (all-request accuracy) submitted100.0% ✎▁····▃▅▃▂▂·█ beats 55/56easy · avg 82%▇ 0.2%noisy
DecisionBench v1 · context management (all-request accuracy) submitted88.0% ✎▁·····▁▇▆▁█▃ beats 38/56easy · avg 73%▇ 0.2%blunt · at its level · noisy
DecisionBench v1 · entity alignment (all-request accuracy) submitted76.4% ✎▁·▁▄▇▄▅▄▂▁▄█ beats 29/56easy · avg 66%▇ 0.1%at its level · noisy
DecisionBench v1 · function agent skill routing (all-request accuracy) submitted74.2% ✎▃▃▄▅▆▂▅▄▃▄█▃ beats 26/56easy · avg 68%▇ 0.1%blunt · at its level · noisy
DecisionBench v1 · routing triage (all-request accuracy) submitted0.0% ✎█▁···▁·▂▆▅▂▂ beats 14/56medium · avg 42%▇ 0.0%blunt · at its level · noisy
consistency█····█·····█ #2 of 3 * · 0.1% of its score · 1 test

Beats Jev 1.13 here, though they rank higher overall. Trails open-jev-deberta-v3-large (DecisionBench 19bf9a64) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Chao Ma · confidence under authority injection0.4120 auroc ✎█····█·····█ beats 1/2avg 0.426 auroc▇ 0.1%blunt · small run

$ where to run it

Download model files ↗
Published model files verified. This is the revision recorded by the benchmark. A local run command and hardware requirements have not been verified for this entry. Source · checked 2026-10-31.

Model repository and documentation: huggingface.co/Mapika ↗

$ how the score was worked out · Hanno Labs DecisionBench v1 r… 95%, Chao Ma · independent injecti… 4%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
Hanno Labs DecisionBench v1 reviewed results · submitted; independent execution unverified27CC0 1.0
Chao Ma · independent injection study2MIT
prior · pull toward the middle
weighted average= 53.0

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json