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.
Together AI's Jev-style classifier. A LoRA fine-tune of Qwen3.5-4B that reads a state, a question and 2 to 24 options and returns one option letter. Served on Together's platform; recipe published.
Decides
choice, score, noul, classify
choice, classify, route
Architecture
hopper
tev
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
Unspecified — weights licence being finalised
Availability
Open weights
Open weights + hosted API
Hosted by
—
Together AI
Input price
—
$0.042/MTok
Decision accuracy
68.5%
88.0%
Calibration error
0.102
—
Valid action rate
—
—
Median latency
—
—
p95 latency
—
—
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 tev1?
hopper is from HopitAI and tev1 from Together AI. hopper has open weights you can download and run; tev1 has open weights and a hosted API. Both answer choice and classify questions. Only hopper answers score and noul. Only tev1 answers route.
Which is more accurate, hopper or tev1?
They report on different suites — hopper 68.5% on JevBench public hard tier (111 items, measured by the authors), tev1 88.0% on Together development set (reused, not held out) — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, hopper or tev1?
hopper: Free (open weights). tev1: $0.042 / $0 per 1M, or free to self-host. Open weights cost nothing per call beyond your own hardware.
Can I run hopper or tev1 locally?
Yes, both: systemone pull hopit-ai/hopper and systemone pull together-ai/tev1 download the weights.
Evaluation suite
JevBench public hard tier (111 items, measured by the authors)
