Contrastive Language Models score a state against a set of candidate actions with a contrastive objective. Two projection heads on frozen Qwen3-8B embeddings, trained with InfoNCE; clm-serve maps Choice, Score and Noul onto candidate ranking.
An open family of System One models. A LoRA adapter plus a pointer head on a frozen Qwen base returns a distribution per typed question in one forward pass, serves TypeSafe's /v1/systemone contract, and ships a fitted temperature with every checkpoint.
Decides
choice, score, noul, rank, classify, route
choice, score, noul, classify, route
Architecture
clm
kev
Fine-tuned from
qwen/qwen3-8b
qwen/qwen3.5-4b-base
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
—
83.8%
Calibration error
—
0.042
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 clm and kev?
clm is from Contrastive-LM and kev from Jared Palmer. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. Only clm answers rank. kev is the smaller model, at 4.0B parameters to 8.0B.
Which is more accurate, clm or kev?
Only kev publishes an accuracy figure (83.8% on transfer-v4 (locked, out of domain)); clm does not, so there is no comparison to make without your own test.
Which is cheaper, clm or kev?
clm: Free (open weights). kev: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run clm or kev locally?
Yes, both: systemone pull contrastive-lm/clm and systemone pull jared-palmer/kev download the weights.
