Models
KTwo models
Both train themselves. Neither is publicly available.
| Model | API ID | Parameters | Modality | Training | Input / MTok | Output / MTok | Availability |
|---|---|---|---|---|---|---|---|
| KTwo Crust 1 Beta New | ktwo-crust-1-beta | 9B | Native multimodal | Self‑training (native) | $0.04 | $0.15 | Private beta |
| KTwo Mantle 0.5 Trained on Hungarian law | Not announced | 30B | Not announced | Self‑training | Not announced | Not announced | Not publicly available |
Not announced yet.
New to the terms in this table? The AI model glossary explains them, and Why 9B? covers what model size means.