Unofficial NoulXP package of Convai Innovations' Laya multilingual (Apache-2.0), fp32-o23: checked against the model's own answers, none changed. Not made by Convai Innovations.
A 4B decision model fine-tuned from Qwen3.5-4B-Base. A state and typed questions (choice, yes/no, score) go in; a temperature-calibrated probability for every option comes out of one pass, through a self-hosted /v1/systemone-compatible server. Apache-2.0 weights and code.
Decides
choice, score, noul, classify, route
choice, score, noul, classify, route
Architecture
laya
wald
Fine-tuned from
convai-innovations/laya
qwen/qwen3.5-4b-base
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
—
88.3%
Calibration error
—
0.045
Valid action rate
—
—
Median latency
44.9 ms
33 ms
p95 latency
742 ms
168 ms
Evaluation suite
Measured by System One Models on 2026-10-05: one question per request, 4 threads of a Xeon Gold 6342
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 laya-multilingual-noulxp and wald?
laya-multilingual-noulxp is from Convai Innovations and wald from ORG2 AI. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. wald reads up to 128K tokens of state, against 1K tokens for laya-multilingual-noulxp. laya-multilingual-noulxp is the smaller model, at 322M parameters to 4.2B.
Which is more accurate, laya-multilingual-noulxp or wald?
Only wald publishes an accuracy figure (88.3% on JevBench public set (231 items; 204 correct), maker's run with the JevBench CLI, v1.2, effort none; items used as a development scoreboard); laya-multilingual-noulxp does not, so there is no comparison to make without your own test.
Which is cheaper, laya-multilingual-noulxp or wald?
laya-multilingual-noulxp: Free (open weights). wald: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Which is faster, laya-multilingual-noulxp or wald?
By their publishers’ figures, wald answers in about 33 ms at the median and laya-multilingual-noulxp in about 44.9 ms — measured on different hardware, so treat it as a rough guide.
Can I run laya-multilingual-noulxp or wald locally?
Yes, both: systemone pull convai-innovations/laya-multilingual-noulxp and systemone pull org2ai/wald download the weights.
JevBench public set (231 items; 204 correct), maker's run with the JevBench CLI, v1.2, effort none; items used as a development scoreboard
Latest version
0.1.0
1.2.0
Variants
noulxp
Dockerfile, LICENSE, NOTICE, model-files, reference, server
