whatmodel.sh / decision / Bespoke Nimble v4 Max (27B)

Bespoke Nimble v4 Max (27B)

+ contribute

Bespoke Nimble v4 Max (27B) is an open-weights decision model from Bespoke Labs, released October 2026. Its published size is 27.36B parameters. Licence: cc-by-nc-4.0.

Catalogue identity notes

Exact Decision Index tested engine nimble-v4-max; reported model Bespoke Nimble v4 Max (27B). Type: full fine-tune; technique: autoregressive. Comparison entry; no current run command, API price or speed inferred from the benchmark. 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 ↗

Publisher parameter count; includes the served backbone, not just adapter size.

70.1±1.4#15 of 355 decision modelsfrom 43 benchmark results across 2 sources

$ how it compares

The decision models ranked either side of Bespoke Nimble v4 Max (27B), out of 357 with a score, then more from Bespoke Labs.

#	model	score	where	RAM	released
13	H2O Lightning 31B v1.1 H2O.ai	70±2	request access
14	Nace.AI Drex Max v1 (27B) Nace-Deepinfra	70±1	model files
15	Bespoke Nimble v4 Max (27B) Bespoke Labs	70±1	model files		2026-10
16	Turion Jev 31B Turion.AI	70±1	model files
17	Perplexity Decider v1.2 (27B) Perplexity	70±1	model files
	more from Bespoke Labs
135	Bespoke Nimble 9B v3 Bespoke Labs	58±1	local	~21GB	2026-10-06
172	Bespoke-Nimble-9B (DecisionBench 594dfdcf) Bespoke Labs	54±2	model files
186	Bespoke-Nimble-9B (JevBench mini) Bespoke Labs	53±2	source code

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

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #12 of 250 · 4.5% of its score · 6 tests

Beats Jeffan-31b-1710, Sieve 31B v3, Fastino GLiDE 1.2 here, though they rank higher overall. Trails Clef-flash, Clef-omni, deck31b (Gemma 4 31B) (decision readout) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: API-Bank (accuracy) submitted	88.0% ✎	▁▁▁▁▁▁▁▂▃▂█▃ beats 218/236	easy · avg 70%	▇ 1.3%	sharp · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	72.9% ✎	▁▁▁▁▁·▁▁▁▄█▃ beats 244/246	medium · avg 60%	▇ 1.0%	blunt · at its level
Decision Index 0.4.1: When2Call (accuracy) submitted	90.7% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 227/246	easy · avg 72%	▇ 0.8%	noisy
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	100.0% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 232/243	medium · avg 37%	▇ 0.8%	sharp · noisy
Decision Index 0.4.1: BFCL (case exact accuracy) submitted	98.0% ✎	▁▂▁▁▁▂▂▄▅▆█▁ beats 232/246	very easy · avg 92%	▇ 0.6%	sharp · maxed out · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	79.6% ✎	no other model yet		—	failed run: Publisher excludes this protocol because its prompt reveals GPT-4 as the best route. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json

retrieval & classification▁▁▁·▁▁▁▁▃▃█▅ #7 of 258 · 7.0% of its score · 7 tests

Beats Fastino GLiDE 1.2, KnowLine 31B Gen1, Ainvest-Decision-27B v1.0 here, though they rank higher overall. Trails StartLux-Decision v1.1 9B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BRIGHT (nDCG@10) submitted	49.7% ✎	▁▁▁▁▁▁▁▂▂▄▇█ beats 226/246	medium · avg 38%	▇ 3.0%	blunt · at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	62.9% ✎	▁▁▁▁▁▁▂▂▃▃█▅ beats 240/246	medium · avg 47%	▇ 2.5%	at its level
Decision Index 0.4.1: HoVer (accuracy) submitted	89.4% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 238/246	easy · avg 74%	▇ 0.5%	blunt
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	91.4% ✎	█▅▅▄▃▃▁▁▁▁·▁ beats 219/245	easy · avg 70%	▇ 0.5%	noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	79.9% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 95/238	easy · avg 79%	▇ 0.3%	noisy
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	91.6% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 151/238	easy · avg 83%	▇ 0.2%	sharp · maxed out · noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	73.3% ✎	no other model yet		—	failed run: Publisher reports a builder bug exposing dialogue turns after the classified turn; awaiting corrected rerun. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json

language understanding▂▂▂▃▂▅▆▃█▆▄▃ #11 of 256 · 18% of its score · 9 tests

Beats Sieve 31B v3, KnowLine 31B Gen1, Ainvest-Decision-27B v1.0 here, though they rank higher overall. Trails Clef-omni, vLLM-SR d3 (27B), AutoTrust-JEV-27B v1.1 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ANLI (macro-F1) submitted	75.5% ✎	▁▂▃▅▅█▂····▁ beats 199/246	medium · avg 60%	▇▇ 4.7%	typical
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	86.1% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 226/246	easy · avg 74%	▇▇ 4.4%	typical
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	61.2% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 191/246	medium · avg 44%	▇ 2.5%	at its level
Decision Index 0.4.1: HellaSwag (accuracy) submitted	97.9% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 236/246	very easy · avg 85%	▇ 2.1%	sharp · maxed out
Decision Index 0.4.1: ACOS (per-review F1) submitted	39.5% ✎	▁··▂▃▃▅▇█▇▃▃ beats 198/246	hard · avg 22%	▇ 1.8%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1) submitted	77.7% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 193/246	medium · avg 62%	▇ 1.0%	typical
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	85.9% ✎	▁▁▁▁▁▁▁▂▂▇▇█ beats 238/246	easy · avg 71%	▇ 0.8%	noisy
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	86.0% ✎	▁▁▁▁▁▁▁▁▁▃▆█ beats 240/246	easy · avg 66%	▇ 0.6%	noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: FinEntity (macro-F1) submitted	94.3% ✎	▁···▁▁▁▁▃▆█▃ beats 221/246	easy · avg 84%	▇ 0.4%	maxed out · noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #31 of 247 · 9.0% of its score · 6 tests

Beats Jeffan-31b-1710, vLLM-SR d3-turbo v1.1 (27B), WeMM-JEV-27B here, though they rank higher overall. Trails Kas 4B, Xor 1.2, Clef-omni here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: New Yorker (accuracy) submitted	81.4% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 228/246	medium · avg 61%	▇ 3.4%	typical
Decision Index 0.4.1: Humicroedit (accuracy) submitted	63.4% ✎	▁·▁▃▂▂▅█▄▂▁▁ beats 181/246	medium · avg 60%	▇ 2.2%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	66.7% ✎	▁········▁▇█ beats 221/246	medium · avg 60%	▇ 1.2%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	80.8% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 222/237	hard · avg 21%	▇ 1.2%	sharp · at its level · noisy
Decision Index 0.4.1: BPoMP (accuracy) submitted	96.6% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 216/246	easy · avg 83%	▇ 0.7%	sharp · maxed out · noisy
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	43.0% ✎	▁▂▄█▇▅▁▁▄▄▁▁ beats 109/246	medium · avg 47%	▇ 0.4%	blunt · at its level

knowledge & reasoning▁▂▂▂▂▃▃▅▅▆█▂ #31 of 263 · 50% of its score · 15 tests

Trails Solar Decide, Clef-omni, Clef here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Etsabary · 1342 text decisions	83.1% ✎	▁··▂·▃▅▁▃▂▂█ beats 18/24	medium · avg 47%	▇▇▇▇▇▇ 14%	sharp
Decision Index 0.4.1: MMLU-Pro (accuracy) submitted	71.0% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 216/246	medium · avg 46%	▇▇▇▇ 10%	sharp · at its level
Decision Index 0.4.1: BBH (accuracy) submitted	93.5% ✎	▄▆▄▄█▇▇▄▃▁▃▁ beats 243/246	easy · avg 66%	▇▇ 4.8%	typical
Decision Index 0.4.1: MMLU (accuracy) submitted	87.9% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 219/246	easy · avg 70%	▇▇ 4.6%	sharp
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	52.5% ✎	▂▆▅▅▆█▆▃▁··▁ beats 216/246	medium · avg 39%	▇▇ 3.9%	blunt · at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	75.9% ✎	▁··▄█▆▆▆▅▅▃▁ beats 208/246	medium · avg 45%	▇ 3.1%	sharp
Decision Index 0.4.1: WinoGrande (accuracy) submitted	94.2% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 228/246	easy · avg 78%	▇ 2.2%	typical
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	23.6% ✎	▁▁▁▁▁▁▁▂▅█▂▁ beats 96/246	hard · avg 22%	▇ 2.1%	at its level

7 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: MuSR (accuracy) submitted	70.0% ✎	▁▁▂█▅▄▁▁▁▁▁▁ beats 215/246	medium · avg 60%	▇ 1.3%	blunt · at its level
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	97.0% ✎	▂▂▂▂▂▂▁▃▄▆█▆ beats 195/246	very easy · avg 90%	▇ 1.2%	sharp · maxed out
Decision Index 0.4.1: CRUXEval (accuracy) submitted	86.7% ✎	██▄▃▃▄▁▁···▁ beats 227/246	medium · avg 58%	▇ 1.0%	noisy
Decision Index 0.4.1: ChessBench (accuracy) submitted	27.3% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 231/246	hard · avg 16%	▇ 1.0%	blunt
Decision Index 0.4.1: CLadder (accuracy) submitted	89.5% ✎	▁███▆▅▃▃▁▁▁▁ beats 220/246	easy · avg 71%	▇ 0.9%	typical
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	99.1% ✎	▁▁▂▁▂▂▂▂▃▆█▅ beats 203/246	very easy · avg 96%	▇ 0.5%	sharp · maxed out · noisy
Decision Index 0.4.1: HLE (accuracy) submitted	12.4% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 121/246	very hard · avg 13%	▇ 0.0%	blunt

workflowsnot ranked on this task yet · 5 tests

test	result	vs the others in its run	counts	why
Foyer inbox · type	45.6% ✎	no other model yet	—	not counted yet: no other model in the same run to compare it with
Foyer inbox · urgency	54.1% ✎	no other model yet	—	not counted yet: no other model in the same run to compare it with
Foyer inbox · needs-reply	74.3% ✎	no other model yet	—	not counted yet: no other model in the same run to compare it with
Foyer inbox · action-48h	53.6% ✎	no other model yet	—	not counted yet: no other model in the same run to compare it with
Foyer inbox · is-noise	85.7% ✎	no other model yet	—	not counted yet: no other model in the same run to compare it with

other tests2 not in a task · 10% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	62.2400 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 224/247	avg 41.206 points	▇▇▇ 6.4%	blunt
Decision Index 0.4.1: private new domains (equated points)	73.0500 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 227/247	avg 38.861 points	▇ 3.5%	blunt

$ where to run it

Download model files ↗
Published model files verified. This is a verified repository snapshot; the benchmark does not establish an identical weight revision. 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/bespokelabs ↗

$ how the score was worked out · Decision Index 0.4.1 85%, Etsabary · decision benchmarks 14%

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

source	results	share of the score	licence
Decision Index 0.4.1 · submitted; independent execution unverified	42	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 85%	CC BY 4.0
Etsabary · decision benchmarks	1	▇▇▇▇ 14%	MIT
prior · pull toward the middle		▇ 1%
weighted average		= 70.1

not counted (7)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	79.6%	failed run: Publisher excludes this protocol because its prompt reveals GPT-4 as the best route. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json
Decision Index 0.4.1	Decision Index 0.4.1: SGD/SGD-X (macro-F1) [correct]	73.3%	failed run: Publisher reports a builder bug exposing dialogue turns after the classified turn; awaiting corrected rerun. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox	Foyer inbox · type [correct]	45.6%	fewer than three models have this test yet, or this source ran it for this model alone
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox	Foyer inbox · urgency [correct]	54.1%	fewer than three models have this test yet, or this source ran it for this model alone
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox	Foyer inbox · needs-reply [correct]	74.3%	fewer than three models have this test yet, or this source ran it for this model alone
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox	Foyer inbox · action-48h [correct]	53.6%	fewer than three models have this test yet, or this source ran it for this model alone
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox	Foyer inbox · is-noise [correct]	85.7%	fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model information Add a place to run it Add a test result Propose a benchmark Correct a benchmark score Edit 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 2026-11-03 02:37 UTC · sources and licences · json