whatmodel.sh / decision / certo-decision-model (DecisionBench 7628c0da)

certo-decision-model (DecisionBench 7628c0da)

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

certo-decision-model (DecisionBench 7628c0da) is an open-weights decision model from altslate. Licence: mit.

Catalogue identity notes

DecisionBench v1 comparison: altslate/certo-decision-model, revision 7628c0da918c5fce96a6629c548630ac6bed8c58; decision-model; adapter jev-compatible-server-certo-custom-head-v1; probability source custom-option-query-softmax-with-score-and-noul-adaptation. 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

· Weights licence ↗

7.4±2.1#355 of 355 decision modelsfrom 27 benchmark results across 1 source

$ how it compares

The decision models ranked either side of certo-decision-model (DecisionBench 7628c0da), out of 357 with a score.

#modelscorewhereRAMreleased
353Nirnaya 0.4B kshitijthakkar16±1model files2026-10-06
354LFM2.5-350M-RLCD notnotsamuel16±1model files
355Zev (decision readout) bhubbard15±1source code
356gte-reranker-modernbert-base (DecisionBench f7481e60) Alibaba7±2model files
357certo-decision-model (DecisionBench 7628c0da) altslate7±2model files

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

$ what it's good at

workflows▃▁▁▂▁▂▂▆█▄▁▁ #112 of 114 · 98% of its score · 27 tests

Beats Supersonic Labs Julia 1, gte-reranker-modernbert-base (DecisionBench f7481e60) here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · action selection (all-request accuracy) submitted13.0% ✎▃▁▂▃▄▃▁▁▂█▇▄ beats 0/56medium · avg 51%▇▇▇▇▇▇ 15%typical
DecisionBench v1 · completion assessment (all-request accuracy) submitted15.0% ✎▃▁▂▃▄·▂▁▁▆█▂ beats 0/56very easy · avg 86%▇▇▇▇▇ 12%sharp
DecisionBench v1 · routing (all-request accuracy) submitted9.7% ✎▃▁▁▁▂▂▁▁▁▁▃█ beats 1/56medium · avg 64%▇▇▇▇ 10%sharp
DecisionBench v1 · verification (all-request accuracy) submitted33.0% ✎▁▃▅▁▃▂▂▁▁▇▇█ beats 2/56medium · avg 56%▇▇▇ 8.4%blunt · at its level
DecisionBench v1 · triage (all-request accuracy) submitted6.0% ✎▃·▁·▂▃▄▃▂█▄▂ beats 0/56easy · avg 78%▇▇▇ 7.8%sharp
DecisionBench v1 · record classification (all-request accuracy) submitted12.0% ✎▃▂▁▂▂▂▄▁▃▆██ beats 3/56medium · avg 39%▇▇ 6.2%typical
DecisionBench v1 · bounded extraction (all-request accuracy) submitted18.1% ✎▃▃▂·▁▁▁▁▁▁█▅ beats 0/56very easy · avg 86%▇▇ 5.5%sharp · noisy
DecisionBench v1 · navigation (all-request accuracy) submitted14.0% ✎▁▁·▃▁▁▁▁▁▁▃█ beats 4/56medium · avg 58%▇▇ 5.3%sharp · too hard for it
19 more tests
testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · content classification (all-request accuracy) submitted9.0% ✎▁····▃▁▁▁▁▁█ beats 2/56easy · avg 73%▇▇ 4.7%typical
DecisionBench v1 · document record classification (all-request accuracy) submitted12.4% ✎▃▂·▁▂▄▃▆▄█▁▁ beats 1/56medium · avg 46%▇▇ 4.4%typical
DecisionBench v1 · rubric scoring prioritization (all-request accuracy) submitted26.5% ✎▁▁▅▅▃▃▅█▃▃·▁ beats 13/56medium · avg 45%▇▇ 3.8%typical
DecisionBench v1 · review scoring (all-request accuracy) submitted17.0% ✎▁······▃▃▅▇█ beats 1/56medium · avg 52%▇ 2.9%typical
DecisionBench v1 · risk scoring (all-request accuracy) submitted23.0% ✎▁▃·▁▂▂▁▃▃█▇▃ beats 2/56medium · avg 62%▇ 2.8%blunt · at its level
DecisionBench v1 · retrieval verification (all-request accuracy) submitted49.5% ✎█▁▂▁▁▁▁▁▂▅▁▁ beats 8/56easy · avg 74%▇ 1.6%at its level · noisy
DecisionBench v1 · document workflows (all-request accuracy) submitted48.2% ✎█▄▃▁▁▃▂▂▁·▁▁ beats 5/56easy · avg 66%▇ 1.3%blunt · at its level
DecisionBench v1 · reasoning (all-request accuracy) submitted33.7% ✎▂·▁▁▆█▅▃▃▂▁▁ beats 10/56medium · avg 52%▇ 1.3%typical
DecisionBench v1 · safety gating (all-request accuracy) submitted47.0% ✎▂·····▆█▂▂▄▇ beats 7/56easy · avg 77%▇ 0.9%at its level · noisy
DecisionBench v1 · guardrails moderation (all-request accuracy) submitted25.0% ✎▁▁▃▂▆▃▃█···▁ beats 4/56medium · avg 52%▇ 0.7%blunt · at its level · noisy
DecisionBench v1 · target selection (all-request accuracy) submitted2.0% ✎▁▁·▃▁▁▃▁▂▁▁█ beats 2/56hard · avg 29%▇ 0.6%too hard for it · noisy
DecisionBench v1 · retrieval ranking (all-request accuracy) submitted3.0% ✎▁▁·▃▁▂▃▂▃▁▁█ beats 4/56medium · avg 35%▇ 0.6%too hard for it · noisy
DecisionBench v1 · semantic filtering (all-request accuracy) submitted55.0% ✎▁·▁▁·▁▁█▂▃█▂ beats 18/56easy · avg 71%▇ 0.6%blunt · at its level · noisy
DecisionBench v1 · content moderation (all-request accuracy) submitted55.0% ✎▁····▃▅▃▂▂·█ beats 14/56easy · avg 82%▇ 0.5%at its level · noisy
DecisionBench v1 · context management (all-request accuracy) submitted41.0% ✎▁·····▃▇▆▁█▃ beats 3/56easy · avg 73%▇ 0.4%blunt · at its level · noisy
DecisionBench v1 · entity alignment (all-request accuracy) submitted14.2% ✎▁·▁▄▇▄▅▄▂▁▄█ beats 7/56easy · avg 66%▇ 0.2%at its level · noisy
DecisionBench v1 · document classification (all-request accuracy) submitted0.0% ✎▃▁·▂▁▁▁▄▁▂▂█ beats 3/56hard · avg 21%▇ 0.2%blunt · noisy
DecisionBench v1 · function agent skill routing (all-request accuracy) submitted13.9% ✎▃▃▄▅▆▂▅▄▃▄█▃ beats 5/56easy · avg 68%▇ 0.2%blunt · at its level · noisy
DecisionBench v1 · routing triage (all-request accuracy) submitted1.4% ✎█▃···▁·▂▆▅▂▂ beats 18/56medium · avg 42%▇ 0.0%blunt · at its level · noisy

$ 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/altslate ↗

$ how the score was worked out · Hanno Labs DecisionBench v1 r… 98%

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
prior · pull toward the middle
weighted average= 7.4

$ 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