A 2.8 MB byte-level transformer for GUI form filling. For each form element it returns one probability per typed option (fill an entity, check, click or skip), using jevlike's option-attention head.
A Situated World Model for fast judgment in agent loops.
Decides
choice
choice, route, classify
Architecture
cua-s1
trio-spark
Fine-tuned from
—
—
License
mit
proprietary
Availability
Open weights
Hosted API
Hosted by
—
MachineFi
Input price
—
$0.042/MTok
Decision accuracy
100.0%
—
Calibration error
—
—
Valid action rate
—
—
Median latency
—
—
p95 latency
—
—
Evaluation suite
196-decision evaluation on real forms
—
Latest version
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 cua-s1-forms and trio-spark-v1?
cua-s1-forms is from Cua and trio-spark-v1 from MachineFi. cua-s1-forms has open weights you can download and run; trio-spark-v1 is only available as a hosted API. Both answer choice questions. Only trio-spark-v1 answers route and classify. cua-s1-forms is licensed mit; trio-spark-v1, proprietary.
Which is more accurate, cua-s1-forms or trio-spark-v1?
Only cua-s1-forms publishes an accuracy figure (100.0% on 196-decision evaluation on real forms); trio-spark-v1 does not, so there is no comparison to make without your own test.
Which is cheaper, cua-s1-forms or trio-spark-v1?
cua-s1-forms: Free (open weights). trio-spark-v1: $0.042 / $0 per 1M. Open weights cost nothing per call beyond your own hardware.
Can I run cua-s1-forms or trio-spark-v1 locally?
cua-s1-forms yes — systemone pull cua/cua-s1-forms downloads its weights. The other is only served as a hosted API.
