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 multimodal decision classifier over text, image, audio and video on Gemma 4 12B, fine-tuned on 30,000 questions. Returns a probability per option and generates nothing.
Decides
choice
choice, noul, score, classify
Architecture
cua-s1
jev-omni
Fine-tuned from
—
google/gemma-4-12b-it
License
mit
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
100.0%
87.6%
Calibration error
—
0.040
Valid action rate
—
—
Median latency
—
83 ms
p95 latency
—
—
Evaluation suite
196-decision evaluation on real forms
DecisionBench Medium (author's set)
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 jev-omni?
cua-s1-forms is from Cua and jev-omni from akhilaaa3. Both have open weights you can download and run. Both answer choice questions. Only jev-omni answers noul, score and classify. cua-s1-forms is the smaller model, at 706K parameters to 12B. cua-s1-forms is licensed mit; jev-omni, apache-2.0.
Which is more accurate, cua-s1-forms or jev-omni?
They report on different suites — cua-s1-forms 100.0% on 196-decision evaluation on real forms, jev-omni 87.6% on DecisionBench Medium (author's set) — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, cua-s1-forms or jev-omni?
cua-s1-forms: Free (open weights). jev-omni: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run cua-s1-forms or jev-omni locally?
Yes, both: systemone pull cua/cua-s1-forms and systemone pull akhilaaa3/jev-omni download the weights.
