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.
Liquid AI's first decision model, served only through the Liquid API. Returns Choice, Score and Noul answers with probability distributions and zero output tokens, on the /v1/systemone request shape used by TypeSafe's SDK.
Decides
choice, score, noul
choice, score, noul, classify, route
Architecture
cygnet
d1
Fine-tuned from
google/gemma-4-12b-it
—
License
mit
proprietary
Availability
Open weights
Hosted API
Hosted by
—
Liquid AI
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 d1?
cygnet is from Blockbrain Labs and d1 from Liquid AI. cygnet has open weights you can download and run; d1 is only available as a hosted API. Both answer choice, score and noul questions. Only d1 answers classify and route. cygnet is licensed mit; d1, proprietary.
Which is more accurate, cygnet or d1?
Only cygnet publishes an accuracy figure (87.9% on JevBench public set (231 items), JevBench CLI); d1 does not, so there is no comparison to make without your own test.
Which is cheaper, cygnet or d1?
cygnet: Free (open weights). d1: Hosted, price not published. Open weights cost nothing per call beyond your own hardware.
Can I run cygnet or d1 locally?
cygnet yes — systemone pull blockbrain-labs/cygnet downloads its weights. The other is only served as a hosted API.
