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.
Shanghai AI Laboratory's multimodal decision model. A fine-tune of Qwen3.5-4B's language backbone (vision tower frozen) that answers up to 16 Choice, Score and Noul questions about a state and up to eight images in one forward pass, with a fitted calibration temperature.
Decides
choice, score, noul
choice, score, noul, classify, route
Architecture
djev
intern-decision
Fine-tuned from
google/diffusiongemma-26b-a4b-it
qwen/qwen3.5-4b
License
apache-2.0
apache-2.0
Availability
Open weights + hosted API
Open weights
Hosted by
Maisa
—
Input price
$0.035/MTok
—
Decision accuracy
—
80.6%
Calibration error
—
—
Valid action rate
—
—
Median latency
—
44 ms
p95 latency
—
44.6 ms
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 djev and intern-decision?
djev is from Maisa and intern-decision from InternLM (Shanghai AI Laboratory). djev has open weights and a hosted API; intern-decision has open weights you can download and run. Both answer choice, score and noul questions. Only intern-decision answers classify and route. intern-decision is the smaller model, at 4.5B parameters to 26B.
Which is more accurate, djev or intern-decision?
Only intern-decision publishes an accuracy figure (80.6% on LocalLLaMA/typed-decisions test split (400 cases, 2,000 decisions)); djev does not, so there is no comparison to make without your own test.
Which is cheaper, djev or intern-decision?
djev: $0.035 / $0 per 1M, or free to self-host. intern-decision: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run djev or intern-decision locally?
Yes, both: systemone pull maisa/djev and systemone pull internlm/intern-decision download the weights.
Evaluation suite
—
LocalLLaMA/typed-decisions test split (400 cases, 2,000 decisions)