The flagship of the vLLM Semantic Router team's Decision 1.0 family. Qwen3.5-9B with a shared candidate head returns a probability for every supplied answer to choice, yes/no and score questions, over a 16,384-token input.
Qwen3.5-4B with a merged rank-16 LoRA distilled from 17,408 teacher questions and 30,000 public training rows, read out as a temperature-scaled softmax over answer-letter logits. Also in 9B, 2B and GGUF.
Decides
choice, score, noul, classify, route
choice, score, noul, classify, route
Architecture
decision
jevk5
Fine-tuned from
qwen/qwen3.5-9b
qwen/qwen3.5-4b
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
77.4%
78.4%
Calibration error
—
0.035
Valid action rate
—
—
Median latency
—
13.2 ms
p95 latency
—
—
Evaluation suite
Figures are from each model’s manifest; accuracy and latency are what the publishers report, on their own suites and hardware. Add a third model.
Questions
What is the difference between decision and jevk5?
decision is from vLLM Semantic Router and jevk5 from Alibi Serikbay. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. jevk5 is the smaller model, at 4.0B parameters to 9.0B.
Which is more accurate, decision or jevk5?
They report on different suites — decision 77.4% on vLLM-SR decision benchmark (54 tasks, 3,766 decisions, weighted), jevk5 78.4% on JevBench public hard tier — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, decision or jevk5?
decision: Free (open weights). jevk5: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run decision or jevk5 locally?
Yes, both: systemone pull vllm-semantic-router/decision and systemone pull alibi-serikbay/jevk5 download the weights.
