whatmodel.sh / decision / vLLM-SR d3-flash (9B)

vLLM-SR d3-flash (9B)

+ contribute

vLLM-SR d3-flash (9B) is an open-weights decision model from vLLM Semantic Router. Its published size is 8.39B parameters. Licence: apache-2.0.

Catalogue identity notes

Exact Decision Index tested engine vllm-sr-d3-flash; reported model vLLM-SR d3-flash (9B). Type: full fine-tune; technique: autoregressive. No 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. Reviewed Ollaya local route added 2026-10-31: d3:flash, manifest 0f5b2d171ede0443170352adff119afc54283be96d50bd6c6cf8e91380ea41a7, matching the independent accuracy-run manifest. Native family implementation; published runtime precision label F32. Memory is measured peak GPU increase (18397 MiB, rounded upward to 18.0 GiB) on RTX 5090 during the short five-question triage workload, not download size or a maximum-context guarantee. Runtime installation is required; no fresh execution performed here. Historical benchmark adapter and input protocol still vary by result.

Model information sources

· Weights licence ↗ · Parameter count ↗

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

60.8±1.2#103 of 355 decision modelsfrom 68 benchmark results across 3 sources

$ how it compares

The decision models ranked either side of vLLM-SR d3-flash (9B), out of 357 with a score, then more from vLLM Semantic Router.

#	model	score	where	RAM	released
101	Verdin-4B v1.1 victorbennett	61±1	model files
102	garas (Gemma 4 31B) plantwaterer	61±1	model files
103	vLLM-SR d3-flash (9B) vLLM Semantic Router	61±1	local	~18GB
104	glm-5.3-flash (DecisionBench openrout) Z.ai	61±2	comparison only		2026-08-26
105	reflex-qwen3.5-4b-lora (DecisionBench 9df1cbbb) kshetrajna12	61±2	model files
	more from vLLM Semantic Router
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
25	vLLM-SR d3 (27B) vLLM Semantic Router	69±1	model files

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

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #91 of 250 · 5.0% of its score · 6 tests

Beats Sieve 27B v2, Jev 1.13, Xor 26B-A4B v1.2 here, though they rank higher overall. Trails OneJev1.6-0.8B, Jev-Omni, baselabs joey-1 (2B) 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	84.8% ✎	▁▁▁▁▁▁▁▂▃▂█▂ beats 152/236	easy · avg 70%	▇ 1.5%	sharp · noisy
Decision Index 0.4.1: BFCL (case exact accuracy) submitted	96.2% ✎	▁▂▁▁▁▂▂▄▅▆█▁ beats 155/246	very easy · avg 92%	▇ 1.0%	sharp · maxed out · noisy
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	53.4% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 136/243	medium · avg 37%	▇ 0.9%	sharp · at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	65.1% ✎	▁▁▁▁▁·▁▁▁▄█▁ beats 138/246	medium · avg 60%	▇ 0.8%	blunt · at its level
Decision Index 0.4.1: When2Call (accuracy) submitted	86.6% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 185/246	easy · avg 72%	▇ 0.8%	noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	79.9% ✎	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▁▁▁·▁▁▁▁▃▃█▅ #115 of 258 · 5.3% of its score · 7 tests

Beats AutoTrust-JEV-27B v1.1, Quyet-1.0-Large, deck31b (Gemma 4 31B) (decision readout) here, though they rank higher overall. Trails vLLM-SR vela3-fast (Vela 3.0, 0.4B), crystal-0.8b, Kas 4B 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	47.3% ✎	▁▁▁▁▁▁▁▂▂▄▇█ beats 190/246	medium · avg 38%	▇ 2.1%	blunt · at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	53.7% ✎	▁▁▁▁▁▁▂▂▃▃█▅ beats 141/246	medium · avg 47%	▇ 1.9%	at its level
Decision Index 0.4.1: HoVer (accuracy) submitted	84.4% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 178/246	easy · avg 74%	▇ 0.5%	blunt
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	68.9% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 124/245	easy · avg 70%	▇ 0.4%	noisy
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	96.4% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 227/238	easy · avg 83%	▇ 0.3%	sharp · noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	92.7% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 225/238	easy · avg 79%	▇ 0.2%	noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	57.9% ✎	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▂▂▂▃▂▅▆▃█▆▄▃ #31 of 256 · 17% of its score · 9 tests

Beats KnowLine 31B Gen1, Ainvest-Decision-27B v1.0, H2O Lightning 31B v1.1 here, though they rank higher overall. Trails 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: ANLI (macro-F1) submitted	70.0% ✎	▁▂▃▅▅█▂····▁ beats 159/246	medium · avg 60%	▇▇ 4.1%	at its level
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	83.6% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 197/246	easy · avg 74%	▇▇ 4.0%	typical
Decision Index 0.4.1: HellaSwag (accuracy) submitted	97.0% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 225/246	very easy · avg 85%	▇ 3.0%	sharp · maxed out
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	51.6% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 143/246	medium · avg 44%	▇ 1.9%	at its level
Decision Index 0.4.1: ACOS (per-review F1) submitted	26.5% ✎	▁··▂▃▃▅▇█▇▃▃ beats 150/246	hard · avg 22%	▇ 1.1%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1) submitted	68.8% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 131/246	medium · avg 62%	▇ 0.8%	at its level
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	73.9% ✎	▁▁▁▁▁▁▁▂▂▇▇█ beats 103/246	easy · avg 71%	▇ 0.8%	noisy
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	76.3% ✎	▁▁▁▁▁▁▁▁▁▃▆█ beats 140/246	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 223/246	easy · avg 84%	▇ 0.4%	noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #65 of 247 · 7.4% of its score · 6 tests

Beats mogan-decision-31B-it, FlexDM · Gemma-4-31B (decision readout), Jev 1.13 here, though they rank higher overall. Trails Crivo-1B, Kas 4B, Aurelis-2B 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	80.3% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 224/246	medium · avg 61%	▇ 3.0%	at its level
Decision Index 0.4.1: Humicroedit (accuracy) submitted	61.0% ✎	▁·▁▃▂▂▅█▄▂▁▁ beats 121/246	medium · avg 60%	▇ 1.6%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	62.9% ✎	▁········▁▇█ beats 142/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	31.9% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 150/237	hard · avg 21%	▇ 0.9%	sharp · at its level · noisy
Decision Index 0.4.1: BPoMP (accuracy) submitted	94.1% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 181/246	easy · avg 83%	▇ 0.8%	sharp · noisy
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	37.4% ✎	▁▂▄█▇▅▁▁▄▄▁▁ beats 34/246	medium · avg 47%	▇ 0.3%	blunt · at its level

knowledge & reasoning▁▂▂▂▂▃▃▅▅▆█▂ #86 of 263 · 38% of its score · 19 tests

Beats StartLux-Decision v1.1 4B, KnowLine 4B Gen5.5, Jeffan-4B-2010 here, though they rank higher overall. Trails Solar Decide, Clef-flash, Xor 1.2 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	64.1% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 172/246	medium · avg 46%	▇▇▇ 8.3%	sharp · at its level
Decision Index 0.4.1: MMLU (accuracy) submitted	83.3% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 168/246	easy · avg 70%	▇▇ 4.9%	sharp
Decision Index 0.4.1: BBH (accuracy) submitted	75.9% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 165/246	easy · avg 66%	▇▇ 4.6%	typical
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	72.8% ✎	▁··▄█▆▆▆▅▅▃▁ beats 196/246	medium · avg 45%	▇ 2.8%	sharp · at its level
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	48.5% ✎	▂▆▅▅▆█▆▃▁··▁ beats 184/246	medium · avg 39%	▇ 2.7%	blunt · at its level
Decision Index 0.4.1: WinoGrande (accuracy) submitted	92.2% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 212/246	easy · avg 78%	▇ 2.2%	typical
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	96.3% ✎	▂▂▂▂▂▂▁▃▄▆█▆ beats 169/246	very easy · avg 90%	▇ 2.0%	sharp · maxed out
Ollaya · PubMedQA	72.8% ✎	▃▂▃▃·▂▃▃▆█▆▂ beats 17/25	easy · avg 69%	▇ 1.7%	typical

11 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	19.6% ✎	▁▁▁▁▁▁▁▂▅█▂▁ beats 70/246	hard · avg 22%	▇ 1.3%	at its level
Ollaya · BoolQ	89.7% ✎	▁··▃▃▁▄▁▃█▇▄ beats 24/25	easy · avg 84%	▇ 1.2%	typical
Decision Index 0.4.1: MuSR (accuracy) submitted	61.0% ✎	▁▁▂█▅▄▁▁▁▁▁▁ beats 143/246	medium · avg 60%	▇ 1.0%	blunt · at its level
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	99.1% ✎	▁▁▂▁▂▂▂▂▃▆█▅ beats 210/246	very easy · avg 96%	▇ 0.9%	sharp · maxed out · noisy
Decision Index 0.4.1: CRUXEval (accuracy) submitted	71.0% ✎	██▄▃▃▄▁▁···▁ beats 171/246	medium · avg 58%	▇ 0.9%	at its level · noisy
Decision Index 0.4.1: CLadder (accuracy) submitted	72.7% ✎	▁███▆▅▃▃▁▁▁▁ beats 143/246	easy · avg 71%	▇ 0.7%	typical
Ollaya · SQuAD 2 answerability	72.2% ✎	▁▁▃▁▇█▃▃▄▃·▁ beats 5/25	easy · avg 80%	▇ 0.7%	blunt
Ollaya · VitaminC verification	83.1% ✎	▁···▃·▄█▃··▁ beats 24/25	easy · avg 78%	▇ 0.6%	blunt
Decision Index 0.4.1: ChessBench (accuracy) submitted	15.1% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 115/246	hard · avg 16%	▇ 0.6%	blunt
Ollaya · MultiNLI	93.3% ✎	▂··▄██▂▆▆█▆▃ beats 25/25	very easy · avg 87%	▇ 0.4%	blunt
Decision Index 0.4.1: HLE (accuracy) submitted	11.4% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 79/246	very hard · avg 13%	▇ 0.0%	blunt

vision▁·▃▂▃▃▇▄▄█▄▃ #35 of 79 · 10% of its score · 12 tests

Beats RSI-Jev v6.1 VL 27B (effort auto), StartLux-Decision v1.1 4B, Surogate Rune 26B-A4B v3 here, though they rank higher overall. Trails Clef-flash, Xor 1.2 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	80.8200 points ✎	▁·▁▁▃▁▄▄▃▅█▂ beats 41/61	avg 63.423 points	▇ 1.3%	blunt
Decision Index Vision 0.4 · MMMU-Pro vision (chance-corrected points) submitted	38.2800 points ✎	▁▁▅▁▄▁▄█▄▃▇▇ beats 40/61	avg 22.797 points	▇ 1.2%	blunt
Decision Index Vision 0.4 · Winoground (chance-corrected points) submitted	82.7500 points ✎	▁·▁▁▁▁·▂▁▃▅█ beats 39/61	avg 66.987 points	▇ 1.1%	blunt
Decision Index Vision 0.4: private same skills (points)	63.9100 points ✎	▁▁▃▃▁▂▄▆▃▃▅█ beats 44/61	avg 48.067 points	▇ 1.1%	blunt
Decision Index Vision 0.4 · Moderation (Hateful Memes) (chance-corrected points) submitted	40.9700 points ✎	▆▃▃▆▄▄▇██▁▆▆ beats 38/61	avg 28.443 points	▇ 1.0%	blunt
Decision Index Vision 0.4 · RealWorldQA (chance-corrected points) submitted	62.2700 points ✎	▁·▂▂▁▃▃▅▅▃▂█ beats 34/61	avg 53.651 points	▇ 0.9%	blunt
Decision Index Vision 0.4 · CharXiv (chance-corrected points) submitted	73.9300 points ✎	▂▁▁▁▁▃▁▅▅▆██ beats 36/61	avg 62.137 points	▇ 0.9%	blunt
Decision Index Vision 0.4 · InfographicVQA (chance-corrected points) submitted	96.1700 points ✎	▁·▁▂▁▁▃▁▂▂▅█ beats 43/61	avg 83.859 points	▇ 0.7%	blunt

4 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index Vision 0.4 · R-Bench-M (chance-corrected points) submitted	20.7700 points ✎	▂▄▇▆█▄▄▇▂··▁ beats 26/61	avg 19.676 points	▇ 0.7%	blunt
Decision Index Vision 0.4 · BLINK (chance-corrected points) submitted	56.0200 points ✎	▁···▁▂▂▂█▄▂▃ beats 38/61	avg 44.994 points	▇ 0.6%	blunt
Decision Index Vision 0.4 · CV-Bench (chance-corrected points) submitted	75.3900 points ✎	▁▁▁▁▂▁▁▁▄█▃▃ beats 32/61	avg 67.662 points	▇ 0.5%	blunt
Decision Index Vision 0.4 · KIE (CORD+FUNSD) (chance-corrected points) submitted	92.9200 points ✎	▁··▁▁▁▁▁▂▄█▁ beats 56/61	avg 85.184 points	▇ 0.4%	blunt

workflows▁▁▁▂▁▂▂▆█▄▁▁ #39 of 114 · 10% of its score · 9 tests

Trails StartLux-Decision v1.1 2B, Solar Decide, simplejev-qwen3.6-35b-a3b (DecisionBench 995ad96e) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Ollaya · Typed decisions test	71.2% ✎	▄·▂▆▄▄▂·▄█▆▆ beats 20/22	medium · avg 64%	▇ 2.8%	at its level
Ollaya · SummEval relevance	45.8% ✎	▁··▁▂▁▅··▃██ beats 14/25	medium · avg 41%	▇ 1.8%	at its level
Ollaya · PAWS paraphrase	88.4% ✎	▂▂▂▂·▃█▆·▃·▁ beats 23/25	easy · avg 81%	▇ 1.4%	typical
Ollaya · Aegis 2 safety	79.6% ✎	▂·▁▁▂·▃▃▂█▃▁ beats 15/25	easy · avg 77%	▇ 1.3%	typical
Ollaya · MASSIVE German intent	88.9% ✎	▁····▁·▂▃▆█▃ beats 21/23	easy · avg 84%	▇ 1.0%	sharp
Ollaya · HelpSteer 2 scoring	45.8% ✎	▁▁··▁·▁▃▇█▅▄ beats 23/23	medium · avg 38%	▇ 0.7%	blunt · at its level
Ollaya · SummEval consistency	84.7% ✎	▁·▁▁··▁▁▁▁▄█ beats 22/25	easy · avg 75%	▇ 0.6%	sharp · noisy
Ollaya · MASSIVE English intent	89.4% ✎	▁·······▂▃█▃ beats 21/23	very easy · avg 86%	▇ 0.6%	noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Ollaya · Civil Comments	92.7% ✎	▂▂▂·▂▅▅▆▃▅·█ beats 19/23	very easy · avg 86%	▇ 0.1%	blunt · noisy

other tests2 not in a task · 7.0% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	54.2100 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 165/247	avg 41.206 points	▇▇ 4.5%	blunt
Decision Index 0.4.1: private new domains (equated points)	54.3600 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 169/247	avg 38.861 points	▇ 2.5%	blunt

$ where to run it

Other ways to run it

Run it as published · needs 18 GB, NVIDIA only · from github.com/ollaya-dev
ollaya run d3:flash --preset triage "The customer requests a refund."

Download model files ↗
Published model files verified. This is a verified repository snapshot; the benchmark does not establish an identical weight revision. A matching Ollaya runtime manifest and published short-workload GPU memory measurement were reviewed separately; see local variants and identity_note. The command was not executed in this review. Source · checked 2026-10-31.

Model repository and documentation: huggingface.co/vllm-sr ↗

$ how the score was worked out · Decision Index 0.4.1 74%, Ollaya · original runs 15%, Decision Index Vision 0.4 10%

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
Ollaya · original runs	14	▇▇▇▇ 15%	Apache-2.0
Decision Index Vision 0.4 · submitted; independent execution unverified	12	▇▇▇ 10%	CC BY 4.0
prior · pull toward the middle		▇ 1%
weighted average		= 60.8

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	79.9%	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]	57.9%	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