whatmodel.sh / decision / vLLM-SR d3-edge (0.6B)

vLLM-SR d3-edge (0.6B)

+ contribute

vLLM-SR d3-edge (0.6B) is an open-weights decision model from vLLM Semantic Router. Its published size is 0.59B parameters. Licence: apache-2.0.

Catalogue identity notes

Exact Decision Index tested engine vllm-sr-d3-edge; reported model vLLM-SR d3-edge (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: d3:edge, manifest 88b3e621e7c659b907938f2a79564734f1f341d0eaace91f9389cdc50555e462, matching the independent accuracy-run manifest. Native family implementation; published runtime precision label F32. Memory is measured peak GPU increase (2657 MiB, rounded upward to 2.6 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.

32.8±0.9#292 of 355 decision modelsfrom 84 benchmark results across 4 sources

$ how it compares

The decision models ranked either side of vLLM-SR d3-edge (0.6B), out of 357 with a score, then more from vLLM Semantic Router.

#	model	score	where	RAM	released
292	Sifr 0.8B v3.1 mohamedlotfy50	33±1	model files
293	jevhome-E (JevBench mini) Jibril Frej	33±2	source code
294	vLLM-SR d3-edge (0.6B) vLLM Semantic Router	33±1	local	~2.6GB
295	Mola-150M codavidgarcia	33±1	model files		2026-10-14
296	DM-JEPA 1.1 Danger Labs	33±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▁▃▁▁▃▂▃▄▆██▆ #214 of 250 · 6.4% of its score · 6 tests

Beats Cam1, InternLM Intern-Decision Qwen3.5-2B, Valen-0.8B here, though they rank higher overall. Trails Prumo (rule-based engine), JevK5 Lite, Crivo-32M v2 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	72.5% ✎	▁▂▁▁▃▂▂▄▅▆█▁ beats 35/246	very easy · avg 92%	▇ 3.7%	sharp · at its level · noisy
Decision Index 0.4.1: API-Bank (accuracy) submitted	40.2% ✎	▁▁▁▁▁▁▁▃▃▂█▂ beats 31/236	easy · avg 70%	▇ 1.4%	sharp · at its level · noisy
Decision Index 0.4.1: When2Call (accuracy) submitted	68.4% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 92/246	easy · avg 72%	▇ 0.7%	at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	57.0% ✎	▁▁▁▁▁·▁▁▁▄█▁ beats 49/246	medium · avg 60%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	0.0% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 16/243	medium · avg 37%	▇ 0.2%	sharp · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	51.6% ✎	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▁▁▁·▁▁▃▁▃▃█▅ #223 of 258 · 11% of its score · 13 tests

Beats Intelif Qwen3-4B, mini-jev Qwen3-4B (decision readout), EXAONE-4.0-1.2B-JEV v0.3 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
System One Bench: MASSIVE intent (Arabic)	55.0% ✎	▄·▁·▁▃▃▃█▄▇▇ beats 2/29	easy · avg 76%	▇ 2.7%	at its level
System One Bench: MIRACL pairwise (Arabic)	76.0% ✎	▂···▃··▃▄▂█▇ beats 2/29	easy · avg 84%	▇ 2.3%	at its level
System One Bench: MIRACL rerank (Arabic)	70.0% ✎	▁▂·▃▁·▂·▄▄█▇ beats 3/29	easy · avg 79%	▇ 1.6%	at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	29.5% ✎	▁▁▁▁▁▁▃▂▃▃█▅ beats 33/246	medium · avg 47%	▇ 0.9%	at its level
Decision Index 0.4.1: BRIGHT (nDCG@10) submitted	31.1% ✎	▁▁▁▁▁▁▁▃▂▄▇█ beats 54/246	medium · avg 38%	▇ 0.9%	blunt · at its level
System One Bench: MASSIVE scenario (Arabic)	41.0% ✎	▃▁·▄▃▂█▂▁▁·▃ beats 2/29	medium · avg 60%	▇ 0.7%	at its level
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	91.9% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 155/238	easy · avg 83%	▇ 0.4%	sharp · at its level · noisy
Decision Index 0.4.1: HoVer (accuracy) submitted	81.0% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 155/246	easy · avg 74%	▇ 0.4%	blunt · at its level

5 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	67.8% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 120/245	easy · avg 70%	▇ 0.3%	at its level · noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	90.0% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 179/238	easy · avg 79%	▇ 0.2%	at its level · noisy
System One Bench: AG News (English)	80.0% ✎	▃···▅▁▅▂▆█▂▅ beats 0/29	easy · avg 88%	▇ 0.2%	blunt
System One Bench: Banking77 (English)	90.0% ✎	█▅▆█▅▃▃▃···▂ beats 16/26	very easy · avg 87%	▇ 0.0%	blunt · noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	51.3% ✎	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▂▂▂▃▂▅▆▃█▆▄▃ #214 of 256 · 22% of its score · 18 tests

Beats Valen-2B, Saina Helm 2 Omni (0.8B), Gevva E2B here, though they rank higher overall. Trails LiquidAI d1-omni-600M, DecisionTune 1.0, vera-spark 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	59.0% ✎	▂▄▂▃▄▇▃▇█▅▃▁ beats 42/246	very easy · avg 85%	▇▇▇ 6.4%	sharp · at its level
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	56.7% ✎	▁▁▁▂▃▃▃▄▇▇█▄ beats 31/246	easy · avg 74%	▇ 3.5%	at its level
Decision Index 0.4.1: ANLI (macro-F1) submitted	40.8% ✎	▁▂▃▅▅█▂····▁ beats 39/246	medium · avg 60%	▇ 2.6%	at its level
System One Bench: BoolQ (English)	79.5% ✎	▁▁▃▃·▁▃▃▇▇█▅ beats 2/29	easy · avg 86%	▇ 1.6%	at its level
System One Bench: OSACT4 offensive (Arabic)	72.5% ✎	▁······▃▃█▇▄ beats 1/29	easy · avg 84%	▇ 1.6%	sharp · at its level
System One Bench: SST-2 (English)	80.5% ✎	▃·▃▁▂▂▃▅█·▂▁ beats 0/29	easy · avg 92%	▇ 1.4%	at its level
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	31.7% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 66/246	medium · avg 44%	▇ 0.8%	at its level
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	56.5% ✎	▁▁▁▁▁▁▁▃▂▇▇█ beats 45/246	easy · avg 71%	▇ 0.6%	at its level · noisy

10 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: FinEntity (macro-F1) submitted	80.5% ✎	▁···▁▁▁▁▃▆█▃ beats 71/246	easy · avg 84%	▇ 0.6%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1) submitted	45.3% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 61/246	medium · avg 62%	▇ 0.5%	at its level
System One Bench: SST-5 (English)	39.5% ✎	▃·▁▁▃·▇▄▄▄█▅ beats 0/29	medium · avg 51%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	35.7% ✎	▁▁▁▁▁▁▁▃▁▃▆█ beats 34/246	easy · avg 66%	▇ 0.4%	at its level · noisy
System One Bench: XNLI French (French)	66.0% ✎	▂▅·▁▁▅▅█▃▁·▃ beats 3/29	medium · avg 76%	▇ 0.3%	blunt · at its level
System One Bench: XNLI Arabic (Arabic)	59.3% ✎	▄▁▃▁▄▇█▅▃▁·▃ beats 5/29	medium · avg 68%	▇ 0.3%	at its level
Decision Index 0.4.1: ACOS (per-review F1) submitted	7.8% ✎	▁··▂▃▃▅▇█▇▃▃ beats 40/246	hard · avg 22%	▇ 0.2%	noisy
System One Bench: Tweet sentiment (Arabic)	61.5% ✎	▁······▁▁▄█▄ beats 5/29	medium · avg 62%	▇ 0.2%	blunt · at its level
System One Bench: XNLI English (English)	84.0% ✎	▃·▁·▃▇█▄█▃▃▁ beats 7/29	easy · avg 85%	▇ 0.1%	blunt · at its level · noisy
System One Bench: DAIR Emotion (English)	57.0% ✎	██▃·····▁··▁ beats 13/29	medium · avg 59%	▇ 0.0%	blunt · at its level

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #208 of 247 · 4.8% of its score · 6 tests

Beats Valen-4B, Saracura PT-BR 4B, TokenRhythm NeoHorse-Jev-4B here, though they rank higher overall. Trails LFM2.5-350M-RLCD, system-one-gemma, Qwen-2.5-1B-RLCD (decision readout) 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	56.6% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 78/246	medium · avg 61%	▇ 2.0%	at its level
Decision Index 0.4.1: BPoMP (accuracy) submitted	54.9% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 39/246	easy · avg 83%	▇ 1.1%	sharp · at its level · noisy
Decision Index 0.4.1: Humicroedit (accuracy) submitted	53.3% ✎	▁·▁▃▃▂▅█▄▂▁▁ beats 45/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	52.1% ✎	▁········▁▇█ beats 35/246	medium · avg 60%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	34.1% ✎	▁▃▄█▇▅▁▁▄▄▁▁ beats 15/246	medium · avg 47%	▇ 0.1%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	1.1% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 18/237	hard · avg 21%	▇ 0.1%	sharp · too hard for it · noisy

knowledge & reasoning▁▂▂▃▂▃▃▅▅▆█▂ #223 of 263 · 36% of its score · 19 tests

Beats Decision-1.0-Eos-0.8B (DecisionBench 3c2d6326), Amber 0.4B, Crivo-400M here, though they rank higher overall. Trails laya (DecisionBench 1c5edc17), LFM2.5-2.6B-RLCD, InternLM Intern-Decision Qwen3.5-0.8B 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	74.6% ✎	▁▁▂▁▃▂▂▂▃▆█▅ beats 35/246	very easy · avg 96%	▇▇▇ 7.0%	sharp · noisy
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	58.0% ✎	▂▂▂▃▂▂▁▃▄▆█▆ beats 38/246	very easy · avg 90%	▇▇▇ 6.9%	sharp · at its level
Decision Index 0.4.1: MMLU (accuracy) submitted	41.2% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 37/246	easy · avg 70%	▇▇ 4.4%	sharp · at its level
Decision Index 0.4.1: BBH (accuracy) submitted	38.3% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 31/246	easy · avg 66%	▇ 3.5%	at its level
Decision Index 0.4.1: MMLU-Pro (accuracy) submitted	23.1% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 46/246	medium · avg 46%	▇ 3.4%	sharp
Decision Index 0.4.1: WinoGrande (accuracy) submitted	61.8% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 50/246	easy · avg 78%	▇ 2.3%	at its level
Ollaya · PubMedQA	48.4% ✎	▃▂▃▃·▂▃▃▆█▆▂ beats 4/25	easy · avg 69%	▇ 1.3%	at its level
Ollaya · BoolQ	72.7% ✎	▁··▃▃▁▄▁▃█▇▄ beats 4/25	easy · avg 84%	▇ 1.2%	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	28.1% ✎	▂▆▅▅▆█▆▃▁··▁ beats 45/246	medium · avg 39%	▇ 1.2%	blunt · at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	34.9% ✎	▁··▄█▆▆▆▅▅▃▁ beats 92/246	medium · avg 45%	▇ 1.1%	sharp
Decision Index 0.4.1: MuSR (accuracy) submitted	49.9% ✎	▁▁▃█▅▄▁▁▁▁▁▁ beats 30/246	medium · avg 60%	▇ 0.6%	blunt · at its level
Decision Index 0.4.1: CLadder (accuracy) submitted	54.1% ✎	▁███▆▅▃▃▁▁▁▁ beats 35/246	easy · avg 71%	▇ 0.6%	at its level
Ollaya · SQuAD 2 answerability	63.9% ✎	▁▁▃▁▇█▃▃▄▃·▁ beats 2/25	easy · avg 80%	▇ 0.6%	blunt · at its level
Ollaya · VitaminC verification	73.8% ✎	▁···▃·▄█▃··▁ beats 6/25	easy · avg 78%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: CRUXEval (accuracy) submitted	43.2% ✎	██▄▃▃▄▁▁···▁ beats 76/246	medium · avg 58%	▇ 0.5%	at its level · noisy
Ollaya · MultiNLI	78.9% ✎	▂··▄██▂▆▆█▆▂ beats 2/25	very easy · avg 87%	▇ 0.4%	blunt
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	11.4% ✎	▁▁▁▁▁▁▃▂▅█▂▁ beats 49/246	hard · avg 22%	▇ 0.3%	typical
Decision Index 0.4.1: ChessBench (accuracy) submitted	7.4% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 11/246	hard · avg 16%	▇ 0.2%	blunt
Decision Index 0.4.1: HLE (accuracy) submitted	16.8% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 228/246	very hard · avg 13%	▇ 0.0%	blunt

vision▃·▃▂▃▃▇▄▄█▄▃ #79 of 79 · 5.6% of its score · 12 tests

Trails LiquidAI d1-omni-600M, InternLM Intern-Decision Qwen3.5-0.8B 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	38.4700 points ✎	▁·▃▁▃▁▄▄▃▅█▂ beats 2/61	avg 63.423 points	▇ 0.7%	blunt
Decision Index Vision 0.4 · MMMU-Pro vision (chance-corrected points) submitted	4.2000 points ✎	▁▃▅▁▄▁▄█▄▃▇▇ beats 1/61	avg 22.797 points	▇ 0.7%	blunt
Decision Index Vision 0.4 · Winoground (chance-corrected points) submitted	27.7500 points ✎	▃·▁▁▁▁·▂▁▃▅█ beats 0/61	avg 66.987 points	▇ 0.6%	blunt
Decision Index Vision 0.4: private same skills (points)	26.8100 points ✎	▃▁▃▃▁▂▄▆▃▃▅█ beats 0/61	avg 48.067 points	▇ 0.6%	blunt
Decision Index Vision 0.4 · Moderation (Hateful Memes) (chance-corrected points) submitted	9.1500 points ✎	▆▃▃▆▄▄▇██▁▆▆ beats 0/61	avg 28.443 points	▇ 0.5%	blunt
Decision Index Vision 0.4 · RealWorldQA (chance-corrected points) submitted	25.6400 points ✎	▃·▂▂▁▃▃▅▅▃▂█ beats 0/61	avg 53.651 points	▇ 0.5%	blunt
Decision Index Vision 0.4 · CharXiv (chance-corrected points) submitted	36.0900 points ✎	▃▁▁▁▁▃▁▅▅▆██ beats 2/61	avg 62.137 points	▇ 0.5%	blunt
Decision Index Vision 0.4 · InfographicVQA (chance-corrected points) submitted	62.8400 points ✎	▃·▁▂▁▁▃▁▂▂▅█ beats 0/61	avg 83.859 points	▇ 0.4%	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	9.4000 points ✎	▃▄▇▆█▄▄▇▂··▁ beats 1/61	avg 19.676 points	▇ 0.4%	blunt
Decision Index Vision 0.4 · BLINK (chance-corrected points) submitted	34.7600 points ✎	▁···▁▃▂▂█▄▂▃ beats 7/61	avg 44.994 points	▇ 0.3%	blunt
Decision Index Vision 0.4 · CV-Bench (chance-corrected points) submitted	54.0200 points ✎	▁▃▁▁▂▁▁▁▄█▃▃ beats 3/61	avg 67.662 points	▇ 0.3%	blunt
Decision Index Vision 0.4 · KIE (CORD+FUNSD) (chance-corrected points) submitted	62.4600 points ✎	▃··▁▁▁▁▁▂▄█▁ beats 0/61	avg 85.184 points	▇ 0.2%	blunt

workflows▁▁▁▂▁▃▂▆█▄▁▁ #90 of 114 · 10% of its score · 10 tests

Beats kev-0.6b (DecisionBench dece6dba), mojev (DecisionBench 0c8695b6), smalljev-semantic-v9 (DecisionBench 8e2bac7c) here, though they rank higher overall. Trails GLiNER2.5-Decide, vLLM-SR Decision 2.0 Kai 0.6B, LiquidAI d1-omni-600M here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Ollaya · Typed decisions test	42.3% ✎	▄·▃▆▄▄▂·▄█▆▆ beats 2/22	medium · avg 64%	▇ 1.9%	at its level
System One Bench: Typed decisions (English)	42.3% ✎	▃▁·▁▁▁▅▄▇██▁ beats 0/29	medium · avg 63%	▇ 1.7%	blunt · at its level
Ollaya · MASSIVE German intent	62.9% ✎	▁····▁·▃▃▆█▃ beats 2/23	easy · avg 84%	▇ 1.6%	sharp · at its level
Ollaya · PAWS paraphrase	82.8% ✎	▂▂▂▂·▃█▆·▂·▁ beats 16/25	easy · avg 81%	▇ 1.5%	at its level
Ollaya · Aegis 2 safety	55.2% ✎	▂·▃▁▂·▃▃▂█▃▁ beats 2/25	easy · avg 77%	▇ 1.2%	at its level
Ollaya · MASSIVE English intent	79.4% ✎	▁·······▂▃█▃ beats 6/23	very easy · avg 86%	▇ 0.8%	at its level · noisy
Ollaya · SummEval relevance	43.8% ✎	▁··▁▂▁▅··▃██ beats 12/25	medium · avg 41%	▇ 0.7%	typical
Ollaya · SummEval consistency	40.3% ✎	▁·▁▁··▁▃▁▁▄█ beats 5/25	easy · avg 75%	▇ 0.7%	sharp · at its level · noisy

2 more tests

test	result	vs the others in its run	difficulty	counts	why
Ollaya · HelpSteer 2 scoring	32.5% ✎	▁▁··▁·▁▃▇█▅▄ beats 5/23	medium · avg 38%	▇ 0.3%	blunt · at its level
Ollaya · Civil Comments	87.7% ✎	▂▂▂·▂▅▅▆▃▅·█ beats 14/23	very easy · avg 86%	▇ 0.1%	blunt · noisy

other tests2 not in a task · 3.8% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	23.5700 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 48/247	avg 41.206 points	▇ 2.4%	blunt
Decision Index 0.4.1: private new domains (equated points)	8.8600 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 14/247	avg 38.861 points	▇ 1.3%	blunt

$ where to run it

Other ways to run it

Run it as published · needs 2.6 GB, NVIDIA only · from github.com/ollaya-dev
ollaya run d3:edge --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 66%, System One Bench (Ali Alhashi… 15%, Ollaya · original runs 13% and 1 more

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	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 66%	CC BY 4.0
System One Bench (Ali Alhashimi)	16	▇▇▇▇▇ 15%	CC-BY-4.0
Ollaya · original runs	14	▇▇▇▇ 13%	Apache-2.0
Decision Index Vision 0.4 · submitted; independent execution unverified	12	▇▇ 6%	CC BY 4.0
prior · pull toward the middle		▇ 0%
weighted average		= 32.8

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	51.6%	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]	51.3%	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

$ 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