Typed decisions from Google's frozen Gemma-4-12B-it, with no fine-tuning. A small shim over unmodified vLLM letters the options, reads the model's own probability for each letter at one answer position and applies one calibration temperature; one output token per decision.
AutoTrust's student of TypeSafe Jev 1.13. A frozen Qwen3.8-27B plus a 108.9M-parameter decision block trained on Jev's own output distributions; one set of weights answers typed questions in one pass (System 1) or generates text with the untouched base (System 2).
Decides
choice, score, noul
choice, score, noul
Architecture
cygnet
blocks-of-experts
Fine-tuned from
google/gemma-4-12b-it
qwen/qwen3.8-27b
License
mit
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
87.9%
88.7%
Calibration error
—
—
Valid action rate
—
—
Median latency
50 ms
137 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 cygnet and jev-27b?
cygnet is from Blockbrain Labs and jev-27b from AutoTrust AI Lab. Both have open weights you can download and run. Both answer choice, score and noul questions. cygnet is the smaller model, at 12B parameters to 27B. cygnet is licensed mit; jev-27b, apache-2.0.
Which is more accurate, cygnet or jev-27b?
They report on different suites — cygnet 87.9% on JevBench public set (231 items), JevBench CLI, jev-27b 88.7% on JevBench public set (231 items), family-macro score, maker's run — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, cygnet or jev-27b?
cygnet: Free (open weights). jev-27b: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Which is faster, cygnet or jev-27b?
By their publishers’ figures, cygnet answers in about 50 ms at the median and jev-27b in about 137 ms — measured on different hardware, so treat it as a rough guide.
Can I run cygnet or jev-27b locally?
Yes, both: systemone pull blockbrain-labs/cygnet and systemone pull autotrust-ai/jev-27b download the weights.
Evaluation suite
JevBench public set (231 items), JevBench CLI
JevBench public set (231 items), family-macro score, maker's run
