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.
Respan's behaviour-scoring model for evals, guardrails and monitoring. For each plain-language behaviour you define, it reads a conversation or agent trace and returns the probability the behaviour is present, absent or not observable, in one forward pass.
Decides
choice, score, noul
noul, classify
Architecture
cygnet
span
Fine-tuned from
google/gemma-4-12b-it
—
License
mit
proprietary
Availability
Open weights
Hosted API
Hosted by
—
Respan
Input price
—
$0.020/MTok
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 span-01?
cygnet is from Blockbrain Labs and span-01 from Respan. cygnet has open weights you can download and run; span-01 is only available as a hosted API. Both answer noul questions. Only cygnet answers choice and score. Only span-01 answers classify. cygnet is licensed mit; span-01, proprietary.
Which is more accurate, cygnet or span-01?
Only cygnet publishes an accuracy figure (87.9% on JevBench public set (231 items), JevBench CLI); span-01 does not, so there is no comparison to make without your own test.
Which is cheaper, cygnet or span-01?
cygnet: Free (open weights). span-01: $0.02 / $0 per 1M. Open weights cost nothing per call beyond your own hardware.
Can I run cygnet or span-01 locally?
cygnet yes — systemone pull blockbrain-labs/cygnet downloads its weights. The other is only served as a hosted API.
