A rank-16 LoRA on Qwen3.5-4B built for the JevBench setting. A document, a policy and a question go in; a calibrated distribution over the option letters comes out of one forward pass. Research and demo use only.
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, score, noul, classify
choice, noul, score, classify
Architecture
hopper
jev-omni
Fine-tuned from
qwen/qwen3.5-4b
google/gemma-4-12b-it
License
Research and demo use only (training data includes RACE, non-commercial); serving code Apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
68.5%
87.6%
Calibration error
0.102
0.040
Valid action rate
—
—
Median latency
—
83 ms
p95 latency
—
—
Evaluation suite
JevBench public hard tier (111 items, measured by the authors)
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 hopper and jev-omni?
hopper is from HopitAI and jev-omni from akhilaaa3. Both have open weights you can download and run. Both answer choice, score, noul and classify questions. hopper is the smaller model, at 4.0B parameters to 12B. hopper is licensed other; jev-omni, apache-2.0.
Which is more accurate, hopper or jev-omni?
They report on different suites — hopper 68.5% on JevBench public hard tier (111 items, measured by the authors), 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, hopper or jev-omni?
hopper: Free (open weights). jev-omni: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run hopper or jev-omni locally?
Yes, both: systemone pull hopit-ai/hopper and systemone pull akhilaaa3/jev-omni download the weights.
