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.
An open System One model from Interfaze AI. A rank-32 LoRA on Qwen3.5-4B reads yes/no, choice and score answers from the logits of one forward pass, returns calibrated probabilities and serves the /v1/systemone API.
Decides
choice, score, noul, classify
choice, score, noul, classify, route
Architecture
hopper
lev
Fine-tuned from
qwen/qwen3.5-4b
qwen/qwen3.5-4b
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%
68.9%
Calibration error
0.102
0.115
Valid action rate
—
—
Median latency
—
69 ms
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 hopper and lev?
hopper is from HopitAI and lev from Interfaze AI. Both have open weights you can download and run. Both answer choice, score, noul and classify questions. Only lev answers route. hopper is licensed other; lev, apache-2.0.
Which is more accurate, hopper or lev?
They report on different suites — hopper 68.5% on JevBench public hard tier (111 items, measured by the authors), lev 68.9% on S1Bench (13 public subsets, macro) — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, hopper or lev?
hopper: Free (open weights). lev: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run hopper or lev locally?
Yes, both: systemone pull hopit-ai/hopper and systemone pull interfaze-ai/lev download the weights.
JevBench public hard tier (111 items, measured by the authors)
