whatmodel.sh / decision / TokenRhythm NeoHorse-Jev-4B

TokenRhythm NeoHorse-Jev-4B

+ contribute

TokenRhythm NeoHorse-Jev-4B is an open-weights decision model from TokenRhythm, released 25 September 2026. Its published size is 4.66B parameters. Licence: apache-2.0.

Catalogue identity notes

Exact Decision Index tested engine neohorse-jev-4b; reported model TokenRhythm NeoHorse-Jev-4B. Type: head / adapter; technique: autoregressive. Implementation: https://github.com/TokenRhythm/NeoHorse/tree/main/jev. 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.

45.7±1.1#239 of 355 decision modelsfrom 56 benchmark results across 3 sources

$ how it compares

The decision models ranked either side of TokenRhythm NeoHorse-Jev-4B, out of 357 with a score.

#	model	score	where	released
239	vera-core-2 scar-ai	46±1	model files
240	Kev-4B (JevBench mini) Jared Palmer	46±2	source code
241	TokenRhythm NeoHorse-Jev-4B TokenRhythm	46±1	model files	2026-09-25
242	SemIf Qwen3.5-4B Base (decision readout) TheoLeeCJ	46±1	source code
243	RSI-Jev v3.0 2B (JevBench mini) Shanghua Gao	45±2	source code

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

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #95 of 250 · 7.6% of its score · 6 tests

Beats Jev 1.13, Xor 26B-A4B v1.2, Senova-4B v2 here, though they rank higher overall. Trails OneJev1.6-0.8B 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	96.2% ✎	▁▂▁▁▁▂▂▄▅▆█▁ beats 158/246	very easy · avg 92%	▇ 2.9%	sharp · noisy
Decision Index 0.4.1: API-Bank (accuracy) submitted	66.1% ✎	▁▁▁▁▁▁▁▂▃▃█▂ beats 83/236	easy · avg 70%	▇ 2.3%	sharp · at its level · noisy
Decision Index 0.4.1: When2Call (accuracy) submitted	51.8% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 46/246	easy · avg 72%	▇ 1.0%	at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	64.5% ✎	▁▁▁▁▁·▁▁▁▄█▁ beats 125/246	medium · avg 60%	▇ 0.8%	blunt · at its level
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	30.7% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 106/243	medium · avg 37%	▇ 0.7%	sharp · at its level · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	79.5% ✎	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▁▁▁·▁▁▁▁▃▃█▅ #177 of 258 · 5.1% of its score · 7 tests

Beats Solar Decide, JevK5 9B v0.3.3, Aplomb 1 here, though they rank higher overall. Trails Mola-68M, Bosun v3.1 1.7B LoRA, Dinah-1 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	40.8% ✎	▁▁▁▁▁▁▁▂▂▄▇█ beats 106/246	medium · avg 38%	▇ 1.7%	blunt · at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	44.4% ✎	▁▁▁▁▁▁▂▂▃▃█▅ beats 80/246	medium · avg 47%	▇ 1.6%	at its level
Decision Index 0.4.1: HoVer (accuracy) submitted	59.3% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 45/246	easy · avg 74%	▇ 0.5%	blunt · at its level
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	53.8% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 63/245	easy · avg 70%	▇ 0.5%	at its level · noisy
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	78.1% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 58/238	easy · avg 83%	▇ 0.4%	sharp · noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	83.8% ✎	▁·▁▁▁▁▁▁▂█▇█ beats 129/238	easy · avg 79%	▇ 0.3%	noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	59.9% ✎	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▂▂▂▃▃▅▆▃█▆▄▃ #201 of 256 · 21% of its score · 9 tests

Trails DecisionTune 1.0, vera-spark, JEV FLAN-T5 Large public v2 (342M) 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	72.2% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 55/246	very easy · avg 85%	▇▇▇ 6.8%	sharp
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	75.0% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 98/246	easy · avg 74%	▇▇ 4.9%	at its level
Decision Index 0.4.1: ANLI (macro-F1) submitted	51.9% ✎	▁▂▃▅▅█▂····▁ beats 74/246	medium · avg 60%	▇▇ 4.3%	at its level
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	38.0% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 83/246	medium · avg 44%	▇ 1.7%	at its level
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	72.6% ✎	▁▁▁▁▁▁▁▂▂▇▇█ beats 90/246	easy · avg 71%	▇ 0.9%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1) submitted	47.4% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 66/246	medium · avg 62%	▇ 0.8%	at its level
Decision Index 0.4.1: FinEntity (macro-F1) submitted	81.9% ✎	▁···▁▁▁▁▃▆█▃ beats 79/246	easy · avg 84%	▇ 0.7%	noisy
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	68.5% ✎	▁▁▁▁▁▁▁▁▁▃▆█ beats 92/246	easy · avg 66%	▇ 0.7%	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	5.4% ✎	▁··▂▃▃▅▇█▇▃▃ beats 26/246	hard · avg 22%	▇ 0.7%	noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #225 of 247 · 7.6% of its score · 6 tests

Trails LFM2.5-350M-RLCD, system-one-gemma, Bekko System One v0 17M 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	52.6% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 64/246	medium · avg 61%	▇ 3.3%	at its level
Decision Index 0.4.1: Humicroedit (accuracy) submitted	51.7% ✎	▁·▁▃▃▂▅█▄▂▁▁ beats 38/246	medium · avg 60%	▇ 1.5%	blunt · at its level
Decision Index 0.4.1: BPoMP (accuracy) submitted	50.5% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 21/246	easy · avg 83%	▇ 1.3%	sharp · noisy
Decision Index 0.4.1: cfcolor (accuracy) submitted	57.2% ✎	▁········▁▇█ beats 71/246	medium · avg 60%	▇ 0.8%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	6.3% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 82/237	hard · avg 21%	▇ 0.4%	sharp · noisy
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	33.8% ✎	▁▃▄█▇▅▁▁▄▄▁▁ beats 14/246	medium · avg 47%	▇ 0.2%	blunt · at its level

knowledge & reasoning▁▂▂▂▂▃▃▅▅▆█▂ #171 of 263 · 41% of its score · 14 tests

Beats Solomon v1.1, Standard One 3B SH (v4), baselabs joey-1 (2B) here, though they rank higher overall. Trails openvons Qwen3-4B (decision readout), Jevstral 8B, Jobe Qwen3.5-4B (decision readout) 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	45.4% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 100/246	medium · avg 46%	▇▇▇ 7.5%	sharp · at its level
Decision Index 0.4.1: MMLU (accuracy) submitted	67.9% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 90/246	easy · avg 70%	▇▇▇ 6.7%	sharp · at its level
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	89.2% ✎	▂▂▂▂▂▂▁▃▄▆█▆ beats 92/246	very easy · avg 90%	▇▇ 6.1%	sharp
Decision Index 0.4.1: BBH (accuracy) submitted	63.7% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 100/246	easy · avg 66%	▇▇ 5.6%	at its level
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	96.3% ✎	▁▁▂▁▂▂▂▂▃▆█▅ beats 94/246	very easy · avg 96%	▇▇ 3.9%	sharp · maxed out · noisy
Decision Index 0.4.1: WinoGrande (accuracy) submitted	68.7% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 72/246	easy · avg 78%	▇ 3.1%	at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	33.7% ✎	▁··▄█▆▆▆▅▅▃▁ beats 87/246	medium · avg 45%	▇ 2.5%	sharp · at its level
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	32.1% ✎	▂▆▅▅▆█▆▃▁··▁ beats 71/246	medium · avg 39%	▇ 2.2%	blunt · at its level

6 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: MuSR (accuracy) submitted	56.1% ✎	▁▁▂█▅▄▁▁▁▁▁▁ beats 83/246	medium · avg 60%	▇ 1.0%	blunt · at its level
Decision Index 0.4.1: CRUXEval (accuracy) submitted	49.6% ✎	██▄▃▃▄▁▁···▁ beats 108/246	medium · avg 58%	▇ 0.9%	at its level · noisy
Decision Index 0.4.1: CLadder (accuracy) submitted	64.9% ✎	▁███▆▅▃▃▁▁▁▁ beats 95/246	easy · avg 71%	▇ 0.8%	at its level
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	21.6% ✎	▁▁▁▁▁▁▁▂▅█▂▁ beats 81/246	hard · avg 22%	▇ 0.8%	typical
Decision Index 0.4.1: ChessBench (accuracy) submitted	11.0% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 75/246	hard · avg 16%	▇ 0.4%	blunt
Decision Index 0.4.1: HLE (accuracy) submitted	10.8% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 60/246	very hard · avg 13%	▇ 0.0%	blunt

vision▁·▃▂▃▃▇▄▄█▄▃ #61 of 79 · 10% of its score · 14 tests

Beats djev diffgemma (decision readout), vLLM-SR d3-nano (2B), StartLux-Decision v1.1 0.8B here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index Vision 0.4 · Mind2Web (chance-corrected points) submitted	73.1400 points ✎	▁·▁▁▃▁▄▄▃▅█▂ beats 28/61	avg 63.423 points	▇ 1.2%	blunt
Decision Index Vision 0.4 · MMMU-Pro vision (chance-corrected points) submitted	25.6800 points ✎	▁▁▅▁▄▁▄█▄▃▇▇ beats 20/61	avg 22.797 points	▇ 1.1%	blunt
Decision Index Vision 0.4 · Winoground (chance-corrected points) submitted	77.5000 points ✎	▁·▁▁▁▁·▂▁▃▅█ beats 25/61	avg 66.987 points	▇ 1.0%	blunt
Decision Index Vision 0.4: private same skills (points)	44.9900 points ✎	▁▁▃▃▁▃▄▆▃▃▅█ beats 14/61	avg 48.067 points	▇ 1.0%	blunt
Decision Index Vision 0.4 · Moderation (Hateful Memes) (chance-corrected points) submitted	20.0300 points ✎	▆▃▃▆▄▄▇██▁▆▆ beats 9/61	avg 28.443 points	▇ 0.8%	blunt
Decision Index Vision 0.4 · RealWorldQA (chance-corrected points) submitted	57.8300 points ✎	▁·▂▂▁▃▃▅▅▃▂█ beats 24/61	avg 53.651 points	▇ 0.8%	blunt
Decision Index Vision 0.4 · CharXiv (chance-corrected points) submitted	64.1600 points ✎	▂▁▁▁▁▃▁▅▅▆██ beats 17/61	avg 62.137 points	▇ 0.8%	blunt
Benchmark Heaven imagejevbench v0.3.0: intelligence (points)	36.8258 points ✎	▃·▃▃··▅█▅▅·▃ beats 3/12	avg 32.235 points	▇ 0.7%	blunt

6 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index Vision 0.4 · InfographicVQA (chance-corrected points) submitted	88.5200 points ✎	▁·▁▂▁▁▃▁▃▂▅█ beats 24/61	avg 83.859 points	▇ 0.6%	blunt
Decision Index Vision 0.4 · R-Bench-M (chance-corrected points) submitted	13.5500 points ✎	▂▄▇▆█▄▄▇▂··▁ beats 7/61	avg 19.676 points	▇ 0.6%	blunt
Decision Index Vision 0.4 · BLINK (chance-corrected points) submitted	-2.1300 points ✎	▃···▁▂▂▂█▄▂▃ beats 2/61	avg 44.994 points	▇ 0.6%	blunt
Decision Index Vision 0.4 · CV-Bench (chance-corrected points) submitted	73.7700 points ✎	▁▁▁▁▂▁▁▁▄█▃▃ beats 26/61	avg 67.662 points	▇ 0.5%	blunt
Decision Index Vision 0.4 · KIE (CORD+FUNSD) (chance-corrected points) submitted	88.9100 points ✎	▁··▁▁▁▁▁▂▄█▁ beats 17/61	avg 85.184 points	▇ 0.3%	blunt
Benchmark Heaven imagejevbench v0.3.0: calibration (points)	78.4171 points ✎	▃···▃·▅█▅▃▃▅ beats 9/12	avg 68.584 points	▇ 0.3%	blunt

other tests2 not in a task · 6.2% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	36.2700 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 83/247	avg 41.206 points	▇▇ 4.0%	blunt
Decision Index 0.4.1: private new domains (equated points)	38.7400 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 118/247	avg 38.861 points	▇ 2.2%	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/TokenRhythm ↗

$ how the score was worked out · Decision Index 0.4.1 89%, Decision Index Vision 0.4 9%, Benchmark Heaven (own decisio… 1%

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	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 89%	CC BY 4.0
Decision Index Vision 0.4 · submitted; independent execution unverified	12	▇▇▇ 9%	CC BY 4.0
Benchmark Heaven (own decision evaluations)	2	▇ 1%	CC BY 4.0
prior · pull toward the middle		▇ 1%
weighted average		= 45.7

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	79.5%	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]	59.9%	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