A training-free layer that turns an open LLM into a decision model. It reads typed choice, yes/no and score answers from one prefill, removes option-order bias with no labels and, from a few hundred labels, calibrates or fits a closed-form head. Weights stay untouched.
A Situated World Model for fast judgment in agent loops.
Decides
choice, score, noul, classify, route
choice, route, classify
Architecture
anyjev
trio-spark
Fine-tuned from
qwen/qwen3-8b
—
License
apache-2.0
proprietary
Availability
Open weights
Hosted API
Hosted by
—
MachineFi
Input price
—
$0.042/MTok
Decision accuracy
77.1%
—
Calibration error
0.034
—
Valid action rate
—
—
Median latency
—
—
p95 latency
—
—
Evaluation suite
LocalLLaMA/typed-decisions (2,000 held-out decisions), L2 with 300 labels per question
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 anyjev and trio-spark-v1?
anyjev is from Nokia and trio-spark-v1 from MachineFi. anyjev has open weights you can download and run; trio-spark-v1 is only available as a hosted API. Both answer choice, classify and route questions. Only anyjev answers score and noul. anyjev is licensed apache-2.0; trio-spark-v1, proprietary.
Which is more accurate, anyjev or trio-spark-v1?
Only anyjev publishes an accuracy figure (77.1% on LocalLLaMA/typed-decisions (2,000 held-out decisions), L2 with 300 labels per question); trio-spark-v1 does not, so there is no comparison to make without your own test.
Which is cheaper, anyjev or trio-spark-v1?
anyjev: Free (open weights). trio-spark-v1: $0.042 / $0 per 1M. Open weights cost nothing per call beyond your own hardware.
Can I run anyjev or trio-spark-v1 locally?
anyjev yes — systemone pull nokia/anyjev downloads its weights. The other is only served as a hosted API.
