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.
Native MLX FP16 conversion of Laya's multilingual checkpoint (mmBERT) for Apple silicon, with a 1,024-token context. The base for multilingual fine-tunes in Laya Studio.
Decides
choice, noul, score, classify
choice, score, noul
Architecture
jev-omni
laya
Fine-tuned from
google/gemma-4-12b-it
convaiinnovations/laya-multilingual
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
87.6%
—
Calibration error
0.040
—
Valid action rate
—
—
Median latency
83 ms
—
p95 latency
—
—
Evaluation suite
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 jev-omni and laya-multilingual-mlx?
jev-omni is from akhilaaa3 and laya-multilingual-mlx from aac6fef. Both have open weights you can download and run. Both answer choice, noul and score questions. Only jev-omni answers classify. laya-multilingual-mlx is the smaller model, at 322M parameters to 12B.
Which is more accurate, jev-omni or laya-multilingual-mlx?
Only jev-omni publishes an accuracy figure (87.6% on DecisionBench Medium (author's set)); laya-multilingual-mlx does not, so there is no comparison to make without your own test.
Which is cheaper, jev-omni or laya-multilingual-mlx?
jev-omni: Free (open weights). laya-multilingual-mlx: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run jev-omni or laya-multilingual-mlx locally?
Yes, both: systemone pull akhilaaa3/jev-omni and systemone pull aac6fef/laya-multilingual-mlx download the weights.
