All models

COMPARE

Put models side by side.

What each one decides, how well calibrated it is, how fast it answers and what it costs to pull. Up to 4 at a time; the better value in each row is marked.

Propertymapika/decider
SummaryThe most-downloaded open System One reproduction. Merged Qwen3.5 fine-tunes, trained on about 95 public decision sets, then calibration-aware RL and a rank-64 LoRA, that softmax letter logits at an answer slot.
Decideschoice, score, noul, classify, route
Architecturedecider
Fine-tuned fromqwen/qwen3.5-2b-base
Licenseapache-2.0
AvailabilityOpen weights
Hosted by—
Input price—
Decision accuracy80.2%
Calibration error0.038
Valid action rate—
Median latency3.2 ms
p95 latency—
Evaluation suiteDecider 67-task regression set
Latest version2026.09
Variants—
Size of latest version3.5 GB
Files4
Downloads0
Stars0
Tagssystem-one, qwen, rl, calibrated, 2b
UpdatedSep 25, 2026