Respan's behaviour-scoring model for evals, guardrails and monitoring. For each plain-language behaviour you define, it reads a conversation or agent trace and returns the probability the behaviour is present, absent or not observable, in one forward pass.
A Situated World Model for fast judgment in agent loops.
Decides
noul, classify
choice, route, classify
Architecture
span
trio-spark
Fine-tuned from
—
—
License
proprietary
proprietary
Availability
Hosted API
Hosted API
Hosted by
Respan
MachineFi
Input price
$0.020/MTok
$0.042/MTok
Decision accuracy
—
—
Calibration error
—
—
Valid action rate
—
—
Median latency
—
—
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 span-01 and trio-spark-v1?
span-01 is from Respan and trio-spark-v1 from MachineFi. Both are only available as a hosted API. Both answer classify questions. Only span-01 answers noul. Only trio-spark-v1 answers choice and route.
Which is more accurate, span-01 or trio-spark-v1?
Neither publishes an accuracy figure. Test both on your own labelled examples.
Which is cheaper, span-01 or trio-spark-v1?
span-01: $0.02 / $0 per 1M. trio-spark-v1: $0.042 / $0 per 1M. Open weights cost nothing per call beyond your own hardware.
