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.
A 0.6B replica of the System One idea trained on game environments. A Qwen3-0.6B backbone with an attention-based Choice head that scores a dynamic candidate set for Maze, Snake, ViZDoom and position prediction, shipped with its full training pipeline.
Decides
choice, score, noul
choice, noul, score
Architecture
cygnet
nanojev
Fine-tuned from
google/gemma-4-12b-it
qwen/qwen3-0.6b
License
mit
mit
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
87.9%
—
Calibration error
—
—
Valid action rate
—
—
Median latency
50 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 nanojev?
cygnet is from Blockbrain Labs and nanojev from TianyuCodings. Both have open weights you can download and run. Both answer choice, score and noul questions. nanojev is the smaller model, at 600M parameters to 12B.
Which is more accurate, cygnet or nanojev?
Only cygnet publishes an accuracy figure (87.9% on JevBench public set (231 items), JevBench CLI); nanojev does not, so there is no comparison to make without your own test.
Which is cheaper, cygnet or nanojev?
cygnet: Free (open weights). nanojev: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run cygnet or nanojev locally?
Yes, both: systemone pull blockbrain-labs/cygnet and systemone pull tianyu-codings/nanojev download the weights.
