Loading comparison…
HEAD TO HEAD
Hanno Labs: bosun and Convai Innovations: laya, compared on what they decide, where they run, what they cost and what their publishers report.
| Property | hanno-labs/bosun | convai-innovations/laya |
|---|---|---|
| Summary | Hanno Labs' small calibrated decision model. A LoRA on Qwen3-1.7B plus trained decision-token embeddings with stable slots, returning the full distribution over up to 255 caller-defined choices and a null slot for Choice, Score and Noul questions. | 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 | bosun | laya |
| Fine-tuned from | qwen/qwen3-1.7b | answerdotai/modernbert-large |
| License | apache-2.0 | apache-2.0 |
| Availability | Open weights | Open weights |
| Hosted by | — | — |
| Input price | — | — |
| Decision accuracy | 84.9% | — |
| Calibration error | 0.050 | — |
| Valid action rate | — | — |
| Median latency | — | 39.5 ms |
| 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.
bosun is from Hanno Labs and laya from Convai Innovations. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. laya is the smaller model, at 421M parameters to 2.0B.
Only bosun publishes an accuracy figure (84.9% on DecisionBench (Hanno-Labs/decision-bench, 23,900 rows; seen task families)); laya does not, so there is no comparison to make without your own test.
bosun: Free (open weights). laya: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Yes, both: systemone pull hanno-labs/bosun and systemone pull convai-innovations/laya download the weights.
| Evaluation suite | DecisionBench (Hanno-Labs/decision-bench, 23,900 rows; seen task families) | — |
|---|
| Latest version | 3.1.0 | 2026.09.25 |
|---|
| Variants | LICENSE, NOTICE, adapter, tokenizer | encoder, eval, multilingual, tokenizer, typed-decisions |
|---|
| Size of latest version | 83.9 MB | 2.2 GB |
|---|
| Files | 17 | 19 |
|---|
| Downloads | 0 | 1 |
|---|
| Stars | 0 | 0 |
|---|
| Tags | system-one, qwen, lora, calibrated, gguf, 2b | system-one, encoder, modernbert, rlcd, multilingual, 421m |
|---|
| Updated | Sep 30, 2026 | Sep 30, 2026 |
|---|