whatmodel.sh / decision / vLLM-SR Decision 2.0 Kai 0.6B

vLLM-SR Decision 2.0 Kai 0.6B

+ contribute

vLLM-SR Decision 2.0 Kai 0.6B is an open-weights decision model from vLLM Semantic Router. Its published size is 0.6B parameters. Licence: apache-2.0.

Catalogue identity notes

Exact Decision Index tested engine decision-2.0-kai-0.6b; reported model vLLM-SR Decision 2.0 Kai 0.6B. 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: decision2:kai, manifest c1f010e05fb8fab53826338931bbd5b5cb7b10ad8410fdf5bf1c344bd6e097f5, matching the independent accuracy-run manifest. Native family implementation; published runtime precision label F32. Memory is measured peak GPU increase (3701 MiB, rounded upward to 3.7 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.

29.0±0.9#315 of 355 decision modelsfrom 56 benchmark results across 2 sources

$ how it compares

The decision models ranked either side of vLLM-SR Decision 2.0 Kai 0.6B, out of 357 with a score, then more from vLLM Semantic Router.

#	model	score	where	RAM	released
315	laya-typed-decisions (DecisionBench bc76315b) ConvAI Innovations	29±2	model files
316	Bosun v3.1 1.7B LoRA Hanno Labs	29±1	model files
317	vLLM-SR Decision 2.0 Kai 0.6B vLLM Semantic Router	29±1	local	~3.7GB
318	Mola-68M codavidgarcia	29±1	model files		2026-10-14
319	JEV Ettin 400M Elfsong	29±1	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▁▃▃▁▁▂▃▄▆██▆ #233 of 250 · 8.4% of its score · 6 tests

Beats DM-JEPA 1.1, JEV FLAN-T5 Large public v2 (342M) here, though they rank higher overall. Trails Qwen-2.5-1B-RLCD (decision readout), Prumo (rule-based engine), GLiNER2.5-Decide here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BFCL (case exact accuracy) submitted	41.6% ✎	▁▂▃▁▁▂▂▄▅▆█▁ beats 17/246	very easy · avg 92%	▇▇ 5.1%	sharp · at its level · noisy
Decision Index 0.4.1: API-Bank (accuracy) submitted	38.8% ✎	▁▁▁▁▁▁▁▃▃▂█▂ beats 28/236	easy · avg 70%	▇ 1.6%	sharp · at its level · noisy
Decision Index 0.4.1: When2Call (accuracy) submitted	38.6% ✎	▁▁▁▃▃▅▅▆█▇█▃ beats 24/246	easy · avg 72%	▇ 0.9%	at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	61.8% ✎	▁▁▁▁▁·▁▁▁▄█▁ beats 86/246	medium · avg 60%	▇ 0.6%	blunt · at its level
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	0.0% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 38/243	medium · avg 37%	▇ 0.1%	sharp · too hard for it · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	71.2% ✎	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▁▁▁·▁▁▃▁▃▃█▅ #228 of 258 · 3.8% of its score · 7 tests

Beats EXAONE-4.0-1.2B-JEV v0.3, Sieve 2B, Strands Decider 2B (Hobson v21) here, though they rank higher overall. Trails Bosun v3.1 0.6B LoRA, Crivo-32M v2, Jev-Style 0.8B v3 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	28.2% ✎	▁▁▁▁▁▁▃▂▃▃█▅ beats 30/246	medium · avg 47%	▇ 1.1%	at its level
Decision Index 0.4.1: BRIGHT (nDCG@10) submitted	34.3% ✎	▁▁▁▁▁▁▁▂▃▄▇█ beats 68/246	medium · avg 38%	▇ 1.1%	blunt · at its level
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	65.2% ✎	▁▁▁▁▁▁▁▃▃▆█▄ beats 38/238	easy · avg 83%	▇ 0.5%	sharp · at its level · noisy
Decision Index 0.4.1: HoVer (accuracy) submitted	53.4% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 19/246	easy · avg 74%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	50.0% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 18/245	easy · avg 70%	▇ 0.4%	at its level · noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	71.6% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 45/238	easy · avg 79%	▇ 0.3%	at its level · noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	59.0% ✎	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▂▂▃▃▂▅▆▃█▆▄▃ #230 of 256 · 20% of its score · 9 tests

Beats vLLM-SR Decision 2.0 Eos 0.8B, Kev 0.8B r15, InternLM Intern-Decision Qwen3.5-0.8B here, though they rank higher overall. Trails JevK5 Lite, LFM2.5-2.6B-RLCD, JEV Ettin 400M here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: HellaSwag (accuracy) submitted	44.5% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 28/246	very easy · avg 85%	▇▇▇ 8.1%	sharp · at its level
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	59.5% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 42/246	easy · avg 74%	▇▇ 4.4%	at its level
Decision Index 0.4.1: ANLI (macro-F1) submitted	30.5% ✎	▁▃▃▅▅█▂····▁ beats 16/246	medium · avg 60%	▇ 3.1%	at its level
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	16.3% ✎	▁·▁·▁▁▃▄▅█▇▄ beats 16/246	medium · avg 44%	▇ 0.9%	typical
Decision Index 0.4.1: FinEntity (macro-F1) submitted	68.0% ✎	▁···▁▁▁▁▃▆█▃ beats 38/246	easy · avg 84%	▇ 0.8%	at its level · noisy
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	31.0% ✎	▁▁▁▁▃▁▁▂▂▇▇█ beats 18/246	easy · avg 71%	▇ 0.8%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1) submitted	34.2% ✎	▁▁▁▃▅▄▄▂▄▆█▄ beats 13/246	medium · avg 62%	▇ 0.6%	at its level
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	1.0% ✎	▁▃▁▁▁▁▁▁▁▃▆█ beats 3/246	easy · avg 66%	▇ 0.5%	at its level · noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ACOS (per-review F1) submitted	4.9% ✎	▁··▂▃▃▅▇█▇▃▃ beats 24/246	hard · avg 22%	▇ 0.3%	noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #235 of 247 · 5.9% of its score · 6 tests

Beats JEV Ettin 150M, JEV FLAN-T5 Large public v2 (342M), Kev 0.8B r15 here, though they rank higher overall. Trails Zev (decision readout), LFM2.5-350M-RLCD, system-one-gemma 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	45.5% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 50/246	medium · avg 61%	▇ 2.4%	at its level
Decision Index 0.4.1: BPoMP (accuracy) submitted	49.2% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 11/246	easy · avg 83%	▇ 1.5%	sharp · at its level · noisy
Decision Index 0.4.1: Humicroedit (accuracy) submitted	49.5% ✎	▁·▁▃▂▂▅█▄▂▁▁ beats 15/246	medium · avg 60%	▇ 1.1%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	51.7% ✎	▁········▁▇█ beats 32/246	medium · avg 60%	▇ 0.6%	blunt · at its level
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	38.0% ✎	▁▂▄█▇▅▁▁▄▄▁▁ beats 39/246	medium · avg 47%	▇ 0.2%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	2.3% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 48/237	hard · avg 21%	▇ 0.1%	sharp · too hard for it · noisy

knowledge & reasoning▁▂▂▃▂▃▃▅▅▆█▂ #228 of 263 · 46% of its score · 19 tests

Beats Decision-1.0-Eos-0.8B (DecisionBench 3c2d6326), Amber 0.4B, vLLM-SR vela3-fast-vision (Vela 3.0, 0.25B) here, though they rank higher overall. Trails laya (DecisionBench 1c5edc17), LFM2.5-2.6B-RLCD, JEV Ettin 400M here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	78.5% ✎	▁▁▂▁▃▂▂▂▃▆█▅ beats 41/246	very easy · avg 96%	▇▇▇▇ 11%	sharp · at its level · noisy
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	53.2% ✎	▂▂▂▃▂▂▁▃▄▆█▆ beats 32/246	very easy · avg 90%	▇▇▇▇ 9.0%	sharp · at its level
Decision Index 0.4.1: MMLU (accuracy) submitted	41.0% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 36/246	easy · avg 70%	▇▇ 5.3%	sharp · at its level
Decision Index 0.4.1: BBH (accuracy) submitted	37.9% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 27/246	easy · avg 66%	▇▇ 4.3%	at its level
Decision Index 0.4.1: MMLU-Pro (accuracy) submitted	18.4% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 35/246	medium · avg 46%	▇ 3.7%	sharp
Decision Index 0.4.1: WinoGrande (accuracy) submitted	57.5% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 38/246	easy · avg 78%	▇ 3.0%	at its level
Ollaya · BoolQ	71.0% ✎	▁··▃▃▁▄▁▃█▇▄ beats 3/25	easy · avg 84%	▇ 1.6%	at its level
Ollaya · PubMedQA	46.8% ✎	▃▂▃▃·▂▃▃▆█▆▂ beats 3/25	easy · avg 69%	▇ 1.6%	at its level

11 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	25.5% ✎	▂▆▅▅▆█▆▃▁··▁ beats 23/246	medium · avg 39%	▇ 1.4%	blunt · at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	25.2% ✎	▁··▄█▆▆▆▅▅▃▁ beats 57/246	medium · avg 45%	▇ 1.2%	sharp
Decision Index 0.4.1: MuSR (accuracy) submitted	44.8% ✎	▁▃▂█▅▄▁▁▁▁▁▁ beats 20/246	medium · avg 60%	▇ 0.7%	blunt · at its level
Ollaya · SQuAD 2 answerability	74.6% ✎	▁▁▃▁▇█▃▃▄▃·▁ beats 9/25	easy · avg 80%	▇ 0.7%	blunt · at its level
Decision Index 0.4.1: CLadder (accuracy) submitted	51.0% ✎	▁███▆▅▃▃▁▁▁▁ beats 20/246	easy · avg 71%	▇ 0.7%	at its level
Ollaya · VitaminC verification	65.1% ✎	▁···▃·▄█▃··▁ beats 2/25	easy · avg 78%	▇ 0.7%	blunt · at its level
Decision Index 0.4.1: CRUXEval (accuracy) submitted	33.2% ✎	██▄▃▃▄▁▁···▁ beats 14/246	medium · avg 58%	▇ 0.6%	at its level · noisy
Ollaya · MultiNLI	80.3% ✎	▂··▄██▂▆▆█▆▂ beats 5/25	very easy · avg 87%	▇ 0.5%	blunt
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	13.7% ✎	▁▁▁▁▁▁▁▃▅█▂▁ beats 53/246	hard · avg 22%	▇ 0.4%	typical
Decision Index 0.4.1: ChessBench (accuracy) submitted	7.1% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 4/246	hard · avg 16%	▇ 0.2%	blunt
Decision Index 0.4.1: HLE (accuracy) submitted	12.0% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 113/246	very hard · avg 13%	▇ 0.0%	blunt

workflows▁▁▁▂▁▃▂▆█▄▁▁ #86 of 114 · 11% of its score · 9 tests

Beats Qwen3-Reranker-4B (DecisionBench 22e68366), kev-0.6b (DecisionBench dece6dba), mojev (DecisionBench 0c8695b6) here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Ollaya · Typed decisions test	47.6% ✎	▄·▂▆▄▄▂·▄█▆▆ beats 5/22	medium · avg 64%	▇ 2.3%	at its level
Ollaya · MASSIVE German intent	74.6% ✎	▁····▁·▂▃▆█▃ beats 6/23	easy · avg 84%	▇ 2.1%	sharp · at its level
Ollaya · PAWS paraphrase	50.4% ✎	▃▂▂▂·▃█▆·▂·▁ beats 0/25	easy · avg 81%	▇ 2.0%	at its level
Ollaya · Aegis 2 safety	62.4% ✎	▂·▁▁▃·▃▃▂█▃▁ beats 4/25	easy · avg 77%	▇ 1.5%	at its level
Ollaya · MASSIVE English intent	78.6% ✎	▁·······▂▃█▃ beats 5/23	very easy · avg 86%	▇ 1.1%	at its level · noisy
Ollaya · SummEval consistency	69.4% ✎	▁·▁▁··▁▁▁▃▄█ beats 7/25	easy · avg 75%	▇ 0.8%	sharp · at its level · noisy
Ollaya · SummEval relevance	20.0% ✎	▁··▁▂▃▅··▃██ beats 4/25	medium · avg 41%	▇ 0.8%	typical
Ollaya · HelpSteer 2 scoring	37.4% ✎	▁▁··▁·▁▃▇█▅▄ beats 11/23	medium · avg 38%	▇ 0.4%	blunt · at its level

1 more test

test	result	vs the others in its run	difficulty	counts	why
Ollaya · Civil Comments	60.0% ✎	▃▂▂·▂▅▅▆▃▅·█ beats 0/23	very easy · avg 86%	▇ 0.1%	blunt · noisy

other tests2 not in a task · 4.7% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	16.0500 points ✎	▂▂▃▃▃▃▄▅▅▇█▃ beats 29/247	avg 41.206 points	▇ 3.0%	blunt
Decision Index 0.4.1: private new domains (equated points)	17.0100 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 64/247	avg 38.861 points	▇ 1.7%	blunt

$ where to run it

Other ways to run it

Run it as published · needs 3.7 GB, NVIDIA only · from github.com/ollaya-dev
ollaya run decision2:kai --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 84%, Ollaya · original runs 16%

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	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 84%	CC BY 4.0
Ollaya · original runs	14	▇▇▇▇▇ 16%	Apache-2.0
prior · pull toward the middle		▇ 0%
weighted average		= 29.0

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	71.2%	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]	59.0%	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