HEAD TO HEAD
Jared Palmer: kev and Convai Innovations: laya, compared on what they decide, where they run, what they cost and what their publishers report.
| Property | jared-palmer/kev | convai-innovations/laya |
|---|---|---|
| Summary | 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. | An open-weights System One model from Convai Innovations. A fully fine-tuned ModernBERT-large encoder with a from-scratch decision head that scores one marker per option and answers every question in a single 33–39 ms pass. Runs on your own CPU or GPU. |
| Decides | choice, score, noul, classify, route | choice, score, noul, classify, route |
| Architecture | kev | laya |
| Fine-tuned from | qwen/qwen3.5-4b-base | answerdotai/modernbert-large |
| 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 | — | 39.5 ms |
| p95 latency | — | — |
| Evaluation suite | transfer-v4 (locked, out of domain) | — |
| Latest version | 2026.09 | 2026.09.25 |
| Variants | — | multilingual, tokenizer, eval, typed-decisions, encoder |
| Size of latest version | 152.2 MB | 2.2 GB |
| Files | 11 | 19 |
| Downloads | 1 | 419 |
| Stars | 0 | 0 |
| Tags | system-one, qwen, lora, pointer-head, 4b | system-one, encoder, modernbert, rlcd, multilingual, 421m |
| Updated | Oct 7, 2026 | Nov 19, 2026 |
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.
kev is from Jared Palmer and laya from Convai Innovations. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. kev reads up to 8K tokens of state, against 512 tokens for laya. laya is the smaller model, at 421M parameters to 4.0B.
Only kev publishes an accuracy figure (83.8% on transfer-v4 (locked, out of domain)); laya does not, so there is no comparison to make without your own test.
kev: Free (open weights). laya: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Yes, both: systemone pull jared-palmer/kev and systemone pull convai-innovations/laya download the weights.