Hanno Labs' small calibrated decision model. A LoRA on Qwen3-1.7B plus trained decision-token embeddings with stable slots, returning the full distribution over up to 255 caller-defined choices and a null slot for Choice, Score and Noul questions.
A typed-decision layer for Google's DiffusionGemma, from David Villalón at Maisa AI. It compiles a request into a small answer canvas, runs one denoising read on patched vLLM and reads the probabilities of the allowed labels, for text, images and images offered as options.
Decides
choice, score, noul, classify, route
choice, score, noul
Architecture
bosun
djev
Fine-tuned from
qwen/qwen3-1.7b
google/diffusiongemma-26b-a4b-it
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights + hosted API
Hosted by
—
Maisa
Input price
—
$0.035/MTok
Decision accuracy
84.9%
—
Calibration error
0.050
—
Valid action rate
—
—
Median latency
—
—
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 bosun and djev?
bosun is from Hanno Labs and djev from Maisa. bosun has open weights you can download and run; djev has open weights and a hosted API. Both answer choice, score and noul questions. Only bosun answers classify and route. bosun is the smaller model, at 2.0B parameters to 26B.
Which is more accurate, bosun or djev?
Only bosun publishes an accuracy figure (84.9% on DecisionBench (Hanno-Labs/decision-bench, 23,900 rows; seen task families)); djev does not, so there is no comparison to make without your own test.
Which is cheaper, bosun or djev?
bosun: Free (open weights). djev: $0.035 / $0 per 1M, or free to self-host. Open weights cost nothing per call beyond your own hardware.
Can I run bosun or djev locally?
Yes, both: systemone pull hanno-labs/bosun and systemone pull maisa/djev download the weights.
Evaluation suite
DecisionBench (Hanno-Labs/decision-bench, 23,900 rows; seen task families)
