An embedding model turned decision model. Qwen3-Embedding-4B with a merged LoRA, driven by the open JevEmbed toolkit, which scores Choice, Score and Noul questions from embedding similarity and returns a distribution without generating text.
Small open typed-decision model with calibrated probabilities, first-class Nepali coverage.
Decides
choice, score, noul, classify
choice, score, noul, classify, route
Architecture
jevembed
mira
Fine-tuned from
qwen/qwen3-embedding-4b
jhu-clsp/mmbert-small
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
85.9%
74.5%
Calibration error
—
0.025
Valid action rate
—
100.0%
Median latency
—
30 ms
p95 latency
—
43 ms
Evaluation suite
JevEmbed-Data test split (64,110 hard-label questions), final LoRA checkpoint before merging
s1-decision-bench
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 jevembed and mira?
jevembed is from HIT-TMG (Lychee Team) and mira from SAGEA. Both have open weights you can download and run. Both answer choice, score, noul and classify questions. Only mira answers route.
Which is more accurate, jevembed or mira?
They report on different suites — jevembed 85.9% on JevEmbed-Data test split (64,110 hard-label questions), final LoRA checkpoint before merging, mira 74.5% on s1-decision-bench — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, jevembed or mira?
jevembed: Free (open weights). mira: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run jevembed or mira locally?
Yes, both: systemone pull hit-tmg/jevembed and systemone pull sagea/mira download the weights.
