whatmodel.sh · find the best decision models you can run

Best Decision models for arts & human taste to run locally

A leaderboard of the best Decision models for arts & human taste you can run on your own computer. Clef-omni is #1 right now, followed by Clef and Jebadiah 27B v2.1 (T1e). Typed-choice results from the Decision Index 0.4.1 public suite. These are decisions over supplied options, not generated answers. Each model shows the download that fits and the command to run it.

25 models · 6 benchmarks · updated

Compare tested decision models and inference recipes across text, vision and workflow tasks. Use task filters and raw results: probability consistency measures coherence, not correctness.

Filter by your hardware and budget · local only · reset
ram[2][4][8][16][24][32][48][64][96][128][192][256]
[]
GB
gpu[apple silicon][graphics card][unified memory][none]
price[local][$0.05][$0.1][any]max per M input tokens
Filter by task
task[all tasks][tools & automation][retrieval & classification][language understanding][arts & human taste][knowledge & reasoning][vision][workflows][consistency]

Scores use reviewed arts & human taste evaluations only. Compare scores within this task. Models without task measurements are excluded.

New #1s:
orRSS

detecting your OS…

$ terminal

detecting your OS…

Run curl "whatmodel.sh/decision?max=0"to get this list in your terminal (add &format=json for agents).

Score vs release date

Release dates available for 9 of 25 models. Undated models are omitted; month-only dates are plotted at the start of the month.

#modeltask score↓where↓RAM↓speed↓get itsources
#modeltask score↓where↓RAM↓speed↓get itsources
22Clef-omni Cloudflare70±1local~87GBhow to run › Decision Index 0.4.1
24Clef Cloudflare69±1†local~68GBhow to run › Decision Index 0.4.1
41Jebadiah 27B v2.1 (T1e) frontier-infra65±1†local~69GBhow to run › Decision Index 0.4.1
44Clef-flash Cloudflare64±1†local~25GBhow to run › Decision Index 0.4.1
65vLLM-SR d3-flash (9B) vLLM Semantic Router61±1†local~18GBhow to run › Decision Index 0.4.1
75Jebadiah 9B v2.1 (T1e) frontier-infra58±1†local~25GBhow to run › Decision Index 0.4.1
83GLiClass v4 edge (0.4B) Knowledgator Engineering57±1†local~2.7GBhow to run › Decision Index 0.4.1
87Perplexity Decider Flash v1.2 (4B) Perplexity56±1†local~10GBhow to run › Decision Index 0.4.1
102Bespoke Nimble 9B v3 Bespoke Labs55±1†local~21GBhow to run › Decision Index 0.4.1
115vLLM-SR d3-mini (4B) vLLM Semantic Router54±1†local~11GBhow to run › Decision Index 0.4.1
133Bespoke Nimble 9B v2 Bespoke Labs50±1†local~20GBhow to run › Decision Index 0.4.1
144Decider 4B Mapika47±1†local~11GBhow to run › Decision Index 0.4.1
157vLLM-SR Decision 2.0 Lux 9B vLLM Semantic Router45±1†local~20GBhow to run › Decision Index 0.4.1
173vLLM-SR Decision 2.0 Nox 4B vLLM Semantic Router40±1†local~12GBhow to run › Decision Index 0.4.1
179vLLM-SR d3-lite (0.8B) vLLM Semantic Router39±1†local~4.5GBhow to run › Decision Index 0.4.1
188Kev 4B r10 Jared Palmer36±1†local~11GBhow to run › Decision Index 0.4.1
192vLLM-SR Decision 2.0 Sol 2B vLLM Semantic Router34±1†local~11GBhow to run › Decision Index 0.4.1
202vLLM-SR d3-nano (2B) vLLM Semantic Router32±1†local~8.2GBhow to run › Decision Index 0.4.1
205Strands Decider 2B (Hobson v21) Amazon Web Services31±1local~6.6GBhow to run › Decision Index 0.4.1
208vLLM-SR d3-edge (0.6B) vLLM Semantic Router31±1†local~2.6GBhow to run › Decision Index 0.4.1
219vLLM-SR Decision 2.0 Eos 0.8B vLLM Semantic Router29±1†local~5.8GBhow to run › Decision Index 0.4.1
235vLLM-SR Decision 2.0 Kai 0.6B vLLM Semantic Router28±1†local~3.7GBhow to run › Decision Index 0.4.1
243Kev 0.8B r15 Jared Palmer27±1†local~4.6GBhow to run › Decision Index 0.4.1
244LiquidAI d1-omni-600M Liquid AI26±1†local~2.7GBhow to run › Decision Index 0.4.1
247CLM-v0.1-8B Qwen3-8B heads Contrastive-LM20±1†local~15GBhow to run › Decision Index 0.4.1

+ 222 more don't fit your machine or budget · show them

247 of 247 catalogue models listed here · sources and licences

* new or thinly tested: ranked on too few benchmark results · † no independently run results verified · ~RAM is estimated; full context may need more memory

25 models · 6 benchmarks · updated