Jev Alternatives: Decision Models with the Same Request Shape
These models accept the same request as Jev, a state plus typed questions, so they can replace it with little or no code change. Models are ranked by their Full score on the Jev Decision Index, a third-party board of 37 benchmarks with private held-out sets, until scores from our own Metatext Decision Benchmark are published.
Ranked list (top 25)
| # | Model | Decision Index | ECE | p50 latency | Per 1k decisions | Params |
|---|---|---|---|---|---|---|
| 1 | vLLM-SR d3 (27B) vLLM Semantic RouterOpen weights | 66.3 | 0.043 | 97 ms | — | 26B |
| 2 | Decider 31B (NVFP4) MapikaOpen weights | 64.6 | 0.024 | 72 ms | — | 33B |
| 3 | Perplexity Decider v1.1 (27B) PerplexityOpen weights | 62.8 | 0.071 | 104 ms | $0.0040 | 28B |
| 4 | Darwin-27B-ZTC FINAL_BenchOpen weights | 62.5 | 0.045 | 104 ms | — | 26B |
| 5 | Fastino GLiDE Fastino | 60.2 | 0.040 | 98 ms | $0.030 | 28B |
| 6 | Torchcast Decision 27B Torchcast AIOpen weights | 59.9 | 0.024 | 100 ms | — | 28B |
| 7 | Shisa DE-2 (repeat) Shisa.AIOpen weights | 59.8 | 0.207 | 57 ms | — | 26B |
| 8 | Matilda Jev v1.5 MaincodeOpen weights | 59.2 | 0.077 | 89 ms | — | 26B |
| 9 | Kev 27B Jared PalmerOpen weights | 58.8 | 0.022 | 135 ms | — | 28B |
| 10 | Quyet-1.0-Large chinhncOpen weights | 58.7 | 0.078 | 105 ms | — | 33B |
| 11 | RSI-Jev v6.1 VL 27B (effort auto) shgaoOpen weights | 58.7 | 0.077 | 76 ms | — | 28B |
| 12 | Wald-26B-A4B ORG2 AIOpen weights | 58.3 | 0.043 | 90 ms | — | 26B |
| 13 | linne-fern 1.0 (27B) MISA-AICOREOpen weights | 57.8 | 0.079 | 108 ms | — | 28B |
| 14 | Blink v0.3 26B-A4B (NVFP4) Pixilab.aiOpen weights | 57.8 | 0.040 | 25 ms | — | 26B |
| 15 | SynACK Decide 26B-A4B v7 SkyPantherOpen weights | 57.5 | 0.069 | 129 ms | — | 26B |
| 16 | Surogate Rune 26B-A4B v3 SurogateOpen weights | 57.4 | 0.120 | 121 ms | — | 26B |
| 17 | Xor 26B-A4B v1.1 Juspay TechnologiesOpen weights | 57.1 | 0.058 | 110 ms | — | 26B |
| 18 | AJev lora1 (Gemma 4 26B-A4B) andyzhang232Open weights | 56.4 | 0.048 | 89 ms | — | 26B |
| 19 | GEV-26B-Decide AutoTrust AI LabOpen weights | 56.2 | 0.072 | 129 ms | — | 26B |
| 20 | Nace.AI Drex v2.0 (35B-A3B) Nace.AIOpen weights | 55.9 | 0.018 | 276 ms | $0.012 | 35B |
| 21 | frontier-infra Jebadiah 27B v2.1 Frontier InfraOpen weights | 55.9 | 0.061 | 110 ms | — | 28B |
| 22 | vLLM-SR Decision 2.0 Vega 27B vLLM Semantic RouterOpen weights | 55.9 | 0.086 | 120 ms | — | 28B |
| 23 | vLLM-SR d3-flash (9B) vLLM Semantic RouterOpen weights | 55.8 | 0.056 | 54 ms | — | 8B |
| 24 | JEV-27B AutoTrust AI LabOpen weights | 55.6 | 0.076 | 102 ms | — | 28B |
| 25 | JADE theunnecessarythingsOpen weights | 55.6 | 0.238 | 133 ms | — | 28B |
Scores, calibration (ECE) and latency from the Jev Decision Index; cost per 1,000 decisions at 200 input tokens each. Updated daily.
More decision-model rankings
Frequently asked questions
What is the best alternative to Jev?
vLLM-SR d3 (27B) is the highest-ranked Jev-compatible model on the Jev Decision Index.
What does Jev-compatible mean?
The model accepts a state and a list of typed questions (choice, score, yes or no) and returns a probability for each allowed answer, the request shape Jev introduced.