An 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.
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, noul, score, classify, route
noul, classify
Architecture
nimble
span
Fine-tuned from
qwen/qwen3.5-9b
—
License
apache-2.0
proprietary
Availability
Open weights + hosted API
Hosted API
Hosted by
Bespoke Labs
Respan
Input price
—
$0.020/MTok
Decision accuracy
90.1%
—
Calibration error
0.054
—
Valid action rate
—
—
Median latency
106 ms
—
p95 latency
—
—
Evaluation suite
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 bespoke-nimble-9b and span-01?
bespoke-nimble-9b is from Bespoke Labs and span-01 from Respan. bespoke-nimble-9b has open weights and a hosted API; span-01 is only available as a hosted API. Both answer noul and classify questions. Only bespoke-nimble-9b answers choice, score and route. bespoke-nimble-9b is licensed apache-2.0; span-01, proprietary.
Which is more accurate, bespoke-nimble-9b or span-01?
Only bespoke-nimble-9b publishes an accuracy figure (90.1% on Bespoke held-out set (324 examples)); span-01 does not, so there is no comparison to make without your own test.
Which is cheaper, bespoke-nimble-9b or span-01?
bespoke-nimble-9b: Hosted, price not published, or free to self-host. span-01: $0.02 / $0 per 1M. Open weights cost nothing per call beyond your own hardware.
Can I run bespoke-nimble-9b or span-01 locally?
bespoke-nimble-9b yes — systemone pull bespoke-labs/bespoke-nimble-9b downloads its weights. The other is only served as a hosted API.
