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.
A 144M-parameter System One model on the multilingual mmBERT-small encoder. Takes a state, a question and 2 to 20 options and returns one decision with probabilities for choice, score and yes/no questions, on a CPU.
Decides
choice, noul, score
choice, score, noul, classify, route
Architecture
bekko
julia
Fine-tuned from
cross-encoder/ettin-reranker-400m-v1
jhu-clsp/mmbert-small
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
—
73.2%
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 julia-1?
bekko-system-one is from Yuichi Tateno (hotchpotch) and julia-1 from Supersonic Labs. Both have open weights you can download and run. Both answer choice, noul and score questions. Only julia-1 answers classify and route. julia-1 is the smaller model, at 144M parameters to 395M. bekko-system-one is licensed other; julia-1, apache-2.0.
Which is more accurate, bekko-system-one or julia-1?
Only julia-1 publishes an accuracy figure (73.2% on typed-decisions test set (400 cases, 2,000 questions)); bekko-system-one does not, so there is no comparison to make without your own test.
Which is cheaper, bekko-system-one or julia-1?
bekko-system-one: Free (open weights). julia-1: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run bekko-system-one or julia-1 locally?
Yes, both: systemone pull hotchpotch/bekko-system-one and systemone pull supersonic-labs/julia-1 download the weights.
5.4 ms
—
Evaluation suite
The maker's short-input speed probe (4-option Choice, RTX 5090, 1 Oct 2026)
typed-decisions test set (400 cases, 2,000 questions)
