whatmodel.sh / decision / StartLux-Decision v1.1 27B

StartLux-Decision v1.1 27B

+ contribute

StartLux-Decision v1.1 27B is an open-weights decision model from StartLux. Its published size is 27.78B parameters. Licence: cc-by-nc-4.0.

Catalogue identity notes

Exact Decision Index tested engine startlux-decision-v1.1-27b-instant; reported model StartLux-Decision v1.1 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

· Weights licence ↗ · Parameter count ↗

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

72.6±1.3#7 of 355 decision modelsfrom 62 benchmark results across 3 sources

$ how it compares

The decision models ranked either side of StartLux-Decision v1.1 27B, out of 357 with a score, then more from StartLux.

#	model	score	where	released
5	KnowLine 31B Gen1 PelaAI	73±2	model files	2026-10-24
6	Ainvest-Decision-27B v1.0 Ainvest-World	73±2	model files	2026-10-27
7	StartLux-Decision v1.1 27B StartLux	73±1	model files
8	vLLM-SR d3-ultra (Decision 3.0, 27B) vLLM Semantic Router	72±1	model files
9	vLLM-SR d3-turbo v1.1 (27B) vLLM Semantic Router	72±1	model files
	more from StartLux
29	StartLux-Decision v1.1 9B StartLux	68±1	model files
54	StartLux-Decision v1.1 4B StartLux	65±1	model files
158	StartLux-Decision v1.1 2B StartLux	55±1	model files

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

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #16 of 250 · 3.7% of its score · 6 tests

Beats Jeffan-31b-1710, Sieve 31B v3, Fastino GLiDE 1.2 here, though they rank higher overall. Trails Solomon v1.1, Clef-flash, Clef-omni 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	84.7% ✎	▁▁▁▁▁▁▁▂▃▂█▂ beats 150/236	easy · avg 70%	▇ 1.1%	sharp · maxed out · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	69.5% ✎	▁▁▁▁▁·▁▁▁▄█▁ beats 229/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: When2Call (accuracy) submitted	93.8% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 246/246	easy · avg 72%	▇ 0.6%	noisy
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	100.0% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 240/243	medium · avg 37%	▇ 0.6%	sharp · noisy
Decision Index 0.4.1: BFCL (case exact accuracy) submitted	97.9% ✎	▁▂▁▁▁▂▂▄▅▆█▁ beats 229/246	very easy · avg 92%	▇ 0.5%	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▁▁▁·▁▁▁▁▃▃█▅ #1 of 258 · 6.3% of its score · 7 tests

Beats Jeffan-31b-1710, baselabs mammoth-1 (27B), Sieve 31B v3 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BRIGHT (nDCG@10) submitted	48.8% ✎	▁▁▁▁▁▁▁▂▂▄▇█ beats 215/246	medium · avg 38%	▇ 2.7%	blunt · at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	65.0% ✎	▁▁▁▁▁▁▂▂▃▃█▅ beats 246/246	medium · avg 47%	▇ 2.3%	at its level
Decision Index 0.4.1: HoVer (accuracy) submitted	90.1% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 243/246	easy · avg 74%	▇ 0.5%	blunt
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	91.8% ✎	█▅▅▄▃▃▁▁▁▁·▁ beats 220/245	easy · avg 70%	▇ 0.4%	noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	92.5% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 223/238	easy · avg 79%	▇ 0.2%	noisy
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	93.8% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 190/238	easy · avg 83%	▇ 0.2%	sharp · maxed out · noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	73.1% ✎	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▂▂▂▃▂▅▆▃█▆▄▃ #4 of 256 · 16% of its score · 9 tests

Beats Jeffan-31b-1710, Sieve 31B v3, Fastino GLiDE 1.2 here, though they rank higher overall. Trails AutoTrust-JEV-27B v1.1, vLLM-SR d3-turbo v1.1 (27B) 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	80.3% ✎	▁▂▃▅▅█▃····▁ beats 240/246	medium · avg 60%	▇▇ 4.1%	typical
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	87.3% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 239/246	easy · avg 74%	▇ 3.7%	typical
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	62.2% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 197/246	medium · avg 44%	▇ 2.2%	at its level
Decision Index 0.4.1: ACOS (per-review F1) submitted	65.7% ✎	▁··▂▃▃▅▇█▇▃▃ beats 244/246	hard · avg 22%	▇ 1.7%	at its level · noisy
Decision Index 0.4.1: HellaSwag (accuracy) submitted	98.1% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 238/246	very easy · avg 85%	▇ 1.6%	sharp · maxed out
Decision Index 0.4.1: VAST (macro-F1) submitted	81.1% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 236/246	medium · avg 62%	▇ 0.8%	typical
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	86.7% ✎	▁▁▁▁▁▁▁▂▂▇▇█ beats 245/246	easy · avg 71%	▇ 0.7%	noisy
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	86.1% ✎	▁▁▁▁▁▁▁▁▁▃▆█ beats 244/246	easy · avg 66%	▇ 0.5%	noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: FinEntity (macro-F1) submitted	95.8% ✎	▁···▁▁▁▁▃▆█▃ beats 242/246	easy · avg 84%	▇ 0.3%	maxed out · noisy

arts & human taste▁▇▄▃▅▅█▅▄▃▁▁ #12 of 247 · 7.9% of its score · 6 tests

Beats Jeffan-31b-1710, KnowLine 31B Gen1 here, though they rank higher overall. Trails AnyJev Tacitus 31B, baselabs elephant-1 (9B), vera-large 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	84.5% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 241/246	medium · avg 61%	▇ 2.9%	typical
Decision Index 0.4.1: Humicroedit (accuracy) submitted	68.8% ✎	▁·▁▃▂▂▅█▄▃▁▁ beats 237/246	medium · avg 60%	▇ 2.0%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	72.1% ✎	▁········▁▇█ beats 245/246	medium · avg 60%	▇ 1.1%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	85.5% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 236/237	hard · avg 21%	▇ 1.0%	sharp · at its level · noisy
Decision Index 0.4.1: BPoMP (accuracy) submitted	98.1% ✎	▁█▄▄▅█▆▅▄▃▁▁ beats 234/246	easy · avg 83%	▇ 0.5%	sharp · maxed out · noisy
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	65.0% ✎	▁▂▄█▇▅▁▁▄▄▁▁ beats 238/246	medium · avg 47%	▇ 0.4%	blunt · at its level

knowledge & reasoning▁▂▂▂▂▃▃▅▅▆█▂ #21 of 263 · 31% of its score · 14 tests

Trails d1 (hosted), Jev 1.13, Matilda Decision v1 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: MMLU-Pro (accuracy) submitted	72.9% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 226/246	medium · avg 46%	▇▇▇ 8.2%	sharp
Decision Index 0.4.1: BBH (accuracy) submitted	91.6% ✎	▄▆▄▄█▇▇▄▃▃▁▁ beats 238/246	easy · avg 66%	▇▇ 4.1%	typical
Decision Index 0.4.1: MMLU (accuracy) submitted	87.5% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 214/246	easy · avg 70%	▇▇ 3.8%	sharp
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	52.0% ✎	▂▆▅▅▆█▆▃▁··▁ beats 211/246	medium · avg 39%	▇ 3.6%	blunt · at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	83.3% ✎	▁··▄█▆▆▆▅▅▃▁ beats 226/246	medium · avg 45%	▇ 2.7%	sharp
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	59.8% ✎	▁▁▁▁▁▁▁▂▅█▂▃ beats 245/246	hard · avg 22%	▇ 2.0%	at its level
Decision Index 0.4.1: WinoGrande (accuracy) submitted	95.1% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 233/246	easy · avg 78%	▇ 1.8%	typical
Decision Index 0.4.1: MuSR (accuracy) submitted	75.3% ✎	▁▁▂█▅▄▃▁▁▁▁▁ beats 231/246	medium · avg 60%	▇ 1.2%	blunt · at its level

6 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ChessBench (accuracy) submitted	25.4% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 225/246	hard · avg 16%	▇ 0.9%	blunt
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	97.4% ✎	▂▂▂▂▂▂▁▃▄▆█▆ beats 222/246	very easy · avg 90%	▇ 0.9%	sharp · maxed out
Decision Index 0.4.1: CRUXEval (accuracy) submitted	86.7% ✎	██▄▃▃▄▁▁···▁ beats 228/246	medium · avg 58%	▇ 0.9%	noisy
Decision Index 0.4.1: CLadder (accuracy) submitted	94.3% ✎	▁███▆▅▃▃▁▃▁▁ beats 236/246	easy · avg 71%	▇ 0.7%	typical
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	99.1% ✎	▁▁▂▁▂▂▂▂▃▆█▅ beats 209/246	very easy · avg 96%	▇ 0.3%	sharp · maxed out · noisy
Decision Index 0.4.1: HLE (accuracy) submitted	17.6% ✎	▁▁▂▃▆█▇▅▅▃▃▁ beats 238/246	very hard · avg 13%	▇ 0.0%	blunt

vision▁·▃▂▃▃▇▄▄█▄▃ #22 of 79 · 15% of its score · 13 tests

Trails Clef-flash, Perplexity Decider Flash v1.2 (4B), Mercury Decide (free endpoint) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index Vision 0.4 · Mind2Web (chance-corrected points) submitted	84.5200 points ✎	▁·▁▁▃▁▄▄▃▅█▂ beats 52/61	avg 63.423 points	▇ 1.7%	blunt
Decision Index Vision 0.4 · MMMU-Pro vision (chance-corrected points) submitted	46.1000 points ✎	▁▁▅▁▄▁▄█▄▃▇▇ beats 55/61	avg 22.797 points	▇ 1.6%	blunt
Decision Index Vision 0.4 · Winoground (chance-corrected points) submitted	86.5000 points ✎	▁·▁▁▁▁·▂▁▃▅█ beats 57/61	avg 66.987 points	▇ 1.5%	blunt
Decision Index Vision 0.4: private same skills (points)	69.3900 points ✎	▁▁▃▃▁▂▄▆▃▃▅█ beats 60/61	avg 48.067 points	▇ 1.5%	blunt
Decision Index Vision 0.4 · Moderation (Hateful Memes) (chance-corrected points) submitted	43.2800 points ✎	▆▃▃▆▄▄▇██▁▆▆ beats 45/61	avg 28.443 points	▇ 1.3%	blunt
JevalKit: VisualDecisionBench images	86.5% ✎	▄▄···█▄·█▄█▄ beats 10/10	easy · avg 78%	▇ 1.2%	blunt
Decision Index Vision 0.4 · RealWorldQA (chance-corrected points) submitted	71.9800 points ✎	▁·▂▂▁▃▃▅▅▃▂█ beats 48/61	avg 53.651 points	▇ 1.2%	blunt
Decision Index Vision 0.4 · CharXiv (chance-corrected points) submitted	79.2000 points ✎	▂▁▁▁▁▃▁▅▅▆██ beats 46/61	avg 62.137 points	▇ 1.1%	blunt

5 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index Vision 0.4 · InfographicVQA (chance-corrected points) submitted	98.3600 points ✎	▁·▁▂▁▁▃▁▂▂▅█ beats 57/61	avg 83.859 points	▇ 0.9%	blunt
Decision Index Vision 0.4 · R-Bench-M (chance-corrected points) submitted	30.1500 points ✎	▂▄▇▆█▄▄▇▂··▁ beats 41/61	avg 19.676 points	▇ 0.9%	blunt
Decision Index Vision 0.4 · BLINK (chance-corrected points) submitted	60.5300 points ✎	▁···▁▂▂▂█▄▂▃ beats 47/61	avg 44.994 points	▇ 0.9%	blunt
Decision Index Vision 0.4 · CV-Bench (chance-corrected points) submitted	76.2000 points ✎	▁▁▁▁▂▁▁▁▄█▃▃ beats 40/61	avg 67.662 points	▇ 0.7%	blunt
Decision Index Vision 0.4 · KIE (CORD+FUNSD) (chance-corrected points) submitted	92.6500 points ✎	▁··▁▁▁▁▁▂▄█▁ beats 51/61	avg 85.184 points	▇ 0.5%	blunt

workflows▁▁▁▂▁▂▂▆█▄▃▁ #2 of 114 · 6.7% of its score · 4 tests

Beats baselabs mammoth-1 (27B) here, though they rank higher overall. Trails Perplexity Decider v1.1 (27B) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
JevalKit: Procedural Typed Decisions	94.2% ✎	█·███▄·▄▄▄·▄ beats 12/12	easy · avg 67%	▇ 3.5%	sharp
JevalKit: DecisionBench standard eval	83.0% ✎	▃▃·▃···▅█▅·█ beats 12/12	easy · avg 70%	▇ 2.0%	typical
JevalKit: JevBench public 231	93.9% ✎	█▄▄▄▄▄██▄··▄ beats 12/12	easy · avg 82%	▇ 1.1%	typical
JevalKit: Typed Decisions held-out modal labels	81.1% ✎	▃▅··▃·▃▃·█▃█ beats 11/12	easy · avg 74%	▇ 0.2%	blunt

other tests5 not in a task · 13% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	64.5100 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 237/247	avg 41.206 points	▇▇ 6.0%	blunt
Decision Index 0.4.1: private new domains (equated points)	73.0200 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 226/247	avg 38.861 points	▇ 3.3%	blunt
JevalKit: JevAlt-Bench EN/TR/DE	94.2% ✎	▅··▃▃▅▃█··▅▃ beats 12/12	very easy · avg 87%	▇ 2.2%	maxed out
JevalKit: Decision Index public 0.3 (official weighted points)	73.0200 points ✎	▃▃·▃·▃▃▃▅▅·█ beats 12/12	avg 39.649 points	▇ 1.4%	blunt
JevalKit: Chinese decision validation v0.2	90.8% ✎	▃····█·█▃▃█▃ beats 9/12	very easy · avg 88%	▇ 0.2%	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/startlux-models ↗

$ how the score was worked out · Decision Index 0.4.1 74%, Decision Index Vision 0.4 14%, JevalKit (Valen Team) 12%

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	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 74%	CC BY 4.0
Decision Index Vision 0.4 · submitted; independent execution unverified	12	▇▇▇▇ 14%	CC BY 4.0
JevalKit (Valen Team)	8	▇▇▇▇ 12%	Apache-2.0
prior · pull toward the middle		▇ 1%
weighted average		= 72.6

not counted (2)

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]	73.1%	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

$ 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