Ultra-small English encoders (17M, 68M and 395M) fine-tuned from Ettin rerankers. Each answers choice, noul and score questions with a probability per candidate and no generated text. Experimental v0 by one developer; the weights have no licence assigned yet.
Ruoxi Qiu's decision adapter for a frozen Qwen3.8-27B: a 6.3M-parameter LoRA on the last four text layers plus a numerical prior. One shared state feeds isolated question branches; each Choice, Noul or Score question returns probabilities with no generated tokens.
Decides
choice, noul, score
choice, noul, score
Architecture
bekko
canopy-jev
Fine-tuned from
cross-encoder/ettin-reranker-400m-v1
qwen/qwen3.8-27b
License
Unassigned: the weights licence is not finalised (treated as non-commercial); training code MIT
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
—
87.4%
Calibration error
—
—
Valid action rate
—
—
Median latency
5 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 bekko-system-one and canopy-jev?
bekko-system-one is from Yuichi Tateno (hotchpotch) and canopy-jev from Ruoxi Qiu. Both have open weights you can download and run. Both answer choice, noul and score questions. bekko-system-one is the smaller model, at 395M parameters to 27B. bekko-system-one is licensed other; canopy-jev, apache-2.0.
Which is more accurate, bekko-system-one or canopy-jev?
Only canopy-jev publishes an accuracy figure (87.4% on JevBench public set (231 items; 202 correct), the maker's own fresh zero-shot run with the upstream scorer); bekko-system-one does not, so there is no comparison to make without your own test.
Which is cheaper, bekko-system-one or canopy-jev?
bekko-system-one: Free (open weights). canopy-jev: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run bekko-system-one or canopy-jev locally?
Yes, both: systemone pull hotchpotch/bekko-system-one and systemone pull camellia86/canopy-jev download the weights.
5.4 ms
—
Evaluation suite
The maker's short-input speed probe (4-option Choice, RTX 5090, 1 Oct 2026)
JevBench public set (231 items; 202 correct), the maker's own fresh zero-shot run with the upstream scorer
