whatmodel.sh / decision / bosun-v3.1-1.7b (DecisionBench 2ef2d972)

bosun-v3.1-1.7b (DecisionBench 2ef2d972)

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

bosun-v3.1-1.7b (DecisionBench 2ef2d972) is an open-weights decision model from Hanno Labs. Its published size is 2.031739904B parameters. Licence: apache-2.0.

Catalogue identity notes

DecisionBench v1 comparison: Hanno-Labs/bosun-v3.1-1.7b, revision 2ef2d97227434a701e8134b90d62ef7669a197d5; decision-model; adapter hf-decision-tokens; probability source native_full_candidate_logits. 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 ↗ · Parameter count ↗

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

66.5±1.9#42 of 355 decision modelsfrom 27 benchmark results across 1 source

$ how it compares

The decision models ranked either side of bosun-v3.1-1.7b (DecisionBench 2ef2d972), out of 357 with a score, then more from Hanno Labs.

#modelscorewhereRAMreleased
40Cerno 27B apertixai67±1model files
41Torchcast Decision 27B Torchcast AI67±1model files2026-10-01
42bosun-v3.1-1.7b (DecisionBench 2ef2d972) Hanno Labs67±2model files
43Decider 31B (NVFP4) Mapika66±1model files2026-10
44Lychee-27B LycheeAI66±1model files
more from Hanno Labs
65bosun-v3.1-0.6b (DecisionBench aaa9dd06) Hanno Labs64±2model files2026-09-22
316Bosun v3.1 1.7B LoRA Hanno Labs29±1model files
330Bosun v3.1 0.6B LoRA Hanno Labs25±1model files2026-09-22

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

$ what it's good at

workflows▁▁▁▂▁▂▂▆█▄▁▁ #7 of 114 · 99% of its score · 27 tests

Beats baselabs mammoth-1 (27B), Nace.AI Drex Max v1 (27B), Perplexity Decider v1.2 (27B) here, though they rank higher overall. Trails Mercury Decide (paid endpoint), Quyet-1.0-Large, StartLux-Decision v1.1 4B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · action selection (all-request accuracy) submitted69.9% ✎▂▁▂▃▄▃▁▁▂█▇▄ beats 56/56medium · avg 51%▇▇▇▇▇▇▇▇▇ 22%at its level
DecisionBench v1 · routing (all-request accuracy) submitted73.3% ✎▁▁▁▁▂▂▁▁▁▁▃█ beats 38/56medium · avg 64%▇▇▇▇▇ 11%sharp
DecisionBench v1 · record classification (all-request accuracy) submitted48.0% ✎▃▂▁▂▂▂▄▁▃▆██ beats 48/56medium · avg 39%▇▇▇▇▇ 11%at its level
DecisionBench v1 · navigation (all-request accuracy) submitted75.0% ✎▁▁·▃▁▁▁▁▁▁▃█ beats 55/56medium · avg 58%▇▇▇▇ 9.4%sharp
DecisionBench v1 · verification (all-request accuracy) submitted64.0% ✎▁▁▅▁▃▂▂▁▁▇▇█ beats 54/56medium · avg 56%▇▇▇ 7.2%blunt · at its level
DecisionBench v1 · document record classification (all-request accuracy) submitted68.4% ✎▂▂·▁▂▄▃▆▄█▁▃ beats 56/56medium · avg 46%▇▇ 6.1%at its level
DecisionBench v1 · rubric scoring prioritization (all-request accuracy) submitted72.8% ✎▁▁▅▅▃▃▅█▃▃·▃ beats 56/56medium · avg 45%▇▇ 5.2%at its level
DecisionBench v1 · review scoring (all-request accuracy) submitted70.0% ✎▁······▂▃▅▇█ beats 56/56medium · avg 52%▇ 3.2%at its level
19 more tests
testresultvs the others in its rundifficultycountswhy
DecisionBench v1 · triage (all-request accuracy) submitted93.0% ✎▁·▁·▂▃▄▃▂█▄▃ beats 55/56easy · avg 78%▇ 3.2%sharp
DecisionBench v1 · target selection (all-request accuracy) submitted42.0% ✎▁▁·▁▁▁▃▁▂▁▁█ beats 38/56hard · avg 29%▇ 2.9%at its level · noisy
DecisionBench v1 · retrieval ranking (all-request accuracy) submitted60.0% ✎▁▁·▁▁▂▃▂▃▁▁█ beats 53/56medium · avg 35%▇ 2.8%at its level · noisy
DecisionBench v1 · content classification (all-request accuracy) submitted80.0% ✎▁····▁▁▁▁▁▁█ beats 50/56easy · avg 73%▇ 2.7%typical
DecisionBench v1 · completion assessment (all-request accuracy) submitted96.0% ✎▁▁▂▃▄·▂▁▁▆█▃ beats 56/56very easy · avg 86%▇ 2.4%sharp · maxed out
DecisionBench v1 · risk scoring (all-request accuracy) submitted75.0% ✎▁▁·▁▂▂▁▃▃█▇▃ beats 56/56medium · avg 62%▇ 2.1%blunt · at its level
DecisionBench v1 · bounded extraction (all-request accuracy) submitted96.3% ✎▃▃▂·▁▁▁▁▁▁█▅ beats 56/56very easy · avg 86%▇ 1.6%sharp · maxed out · noisy
DecisionBench v1 · reasoning (all-request accuracy) submitted39.7% ✎▂·▁▁▆█▅▃▃▂▁▁ beats 17/56medium · avg 52%▇ 1.4%at its level
DecisionBench v1 · document workflows (all-request accuracy) submitted93.7% ✎█▄▃▁▁▃▂▂▁·▁▃ beats 56/56easy · avg 66%▇ 0.8%blunt · at its level
DecisionBench v1 · retrieval verification (all-request accuracy) submitted92.1% ✎█▁▂▁▁▁▁▁▂▅▁▃ beats 56/56easy · avg 74%▇ 0.7%noisy
DecisionBench v1 · guardrails moderation (all-request accuracy) submitted90.2% ✎▁▁▁▂▆▃▃█···▃ beats 56/56medium · avg 52%▇ 0.6%blunt · at its level · noisy
DecisionBench v1 · document classification (all-request accuracy) submitted32.0% ✎▂▁·▂▁▁▁▄▁▂▂█ beats 39/56hard · avg 21%▇ 0.5%blunt · at its level · noisy
DecisionBench v1 · safety gating (all-request accuracy) submitted96.0% ✎▂·····▆█▂▂▄▇ beats 45/56easy · avg 77%▇ 0.4%noisy
DecisionBench v1 · semantic filtering (all-request accuracy) submitted88.0% ✎▁·▁▁·▁▁█▂▃█▃ beats 53/56easy · avg 71%▇ 0.3%blunt · noisy
DecisionBench v1 · context management (all-request accuracy) submitted96.0% ✎▁·····▁▇▆▁█▃ beats 56/56easy · avg 73%▇ 0.2%blunt · noisy
DecisionBench v1 · content moderation (all-request accuracy) submitted99.0% ✎▁····▃▅▃▂▂·█ beats 43/56easy · avg 82%▇ 0.2%noisy
DecisionBench v1 · entity alignment (all-request accuracy) submitted99.9% ✎▁·▁▄▇▄▅▄▂▁▄█ beats 54/56easy · avg 66%▇ 0.2%noisy
DecisionBench v1 · function agent skill routing (all-request accuracy) submitted98.1% ✎▃▃▄▅▆▂▅▄▃▄█▃ beats 52/56easy · avg 68%▇ 0.1%blunt · noisy
DecisionBench v1 · routing triage (all-request accuracy) submitted90.9% ✎█▁···▁·▂▆▅▂▃ beats 55/56medium · avg 42%▇ 0.0%blunt · at its level · noisy

$ where to run it

Download model files ↗
Published model files verified. Includes adapters or readout weights; follow the model documentation for the required base model. 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/Hanno-Labs ↗

$ how the score was worked out · Hanno Labs DecisionBench v1 r… 99%

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= 66.5

$ 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