Omar (kouhxp)'s CPU decision model: a Qwen3.5-0.8B fine-tune shipped as GGUF for llama.cpp that answers yes/no, choice and score questions with a probability per option plus a 'none of these' reject probability, served by a local Jev-style HTTP runtime.
An independent Jev-shaped reproduction on DeBERTa-v3-large. One state and any number of choice, score and yes/no questions go in, and a calibrated distribution per question comes out of one forward pass. Public gold labels only.
Decides
choice, score, noul, classify, route
choice, score, noul, classify
Architecture
gutsy
open-jev
Fine-tuned from
qwen/qwen3.5-0.8b
microsoft/deberta-v3-large
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
73.2%
85.4%
Calibration error
—
0.022
Valid action rate
—
—
Median latency
—
28 ms
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 gutsy and open-jev-deberta-v3-large?
gutsy is from Omar (kouhxp) and open-jev-deberta-v3-large from Kotoba Labs. Both have open weights you can download and run. Both answer choice, score, noul and classify questions. Only gutsy answers route. gutsy reads up to 8K tokens of state, against 512 tokens for open-jev-deberta-v3-large. open-jev-deberta-v3-large is the smaller model, at 434M parameters to 800M.
Which is more accurate, gutsy or open-jev-deberta-v3-large?
They report on different suites — gutsy 73.2% on JevBench public set (231 items; 169 correct), the maker's own run with Q8_0, open-jev-deberta-v3-large 85.4% on Kotoba held-out test, seen question types (banking77, SST-5, BoolQ) — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, gutsy or open-jev-deberta-v3-large?
gutsy: Free (open weights). open-jev-deberta-v3-large: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run gutsy or open-jev-deberta-v3-large locally?
Yes, both: systemone pull kouhxp/gutsy and systemone pull kotoba-labs/open-jev-deberta-v3-large download the weights.
Evaluation suite
JevBench public set (231 items; 169 correct), the maker's own run with Q8_0
Kotoba held-out test, seen question types (banking77, SST-5, BoolQ)
