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.

Propertybespoke-labs/bespoke-nimble-9b
SummaryAn open Jev-style LoRA on Qwen3.5-9B from Bespoke Labs, trained on 2,676 contrastively curated examples to score the allowed answer tokens directly for enums, booleans and rubric levels. Recipe, data and a public benchmark suite are released with it.
Decideschoice, noul, score, classify, route
Architecturenimble
Fine-tuned fromqwen/qwen3.5-9b
Licenseapache-2.0
AvailabilityOpen weights + hosted API
Hosted byBespoke Labs
Input price—
Decision accuracy90.1%
Calibration error0.054
Valid action rate—
Median latency106 ms
p95 latency—
Evaluation suiteBespoke held-out set (324 examples)
Latest version2026.09
VariantsSHA256SUMS
Size of latest version184.3 MB
Files10
Downloads0
Stars0
Tagssystem-one, qwen, lora, curated-data, 9b
UpdatedSep 25, 2026