Aion 1.0 Instruct 2.3B vs Hy4 flash Base
Compare documented capabilities and access. A dash means a comparable figure is unavailable.
| Specification | ||
|---|---|---|
| Task | On-device language generation | Multimodal pretraining |
| Published capabilities | Windows-oriented instruction model for local text generation, including document and email tasks. | Pretrained checkpoint for continued training and adaptation. The Flash architecture uses a sparse Engram memory module. |
| Model size | 2.3B parameters | Approximately 220B total / 12B active per token |
| Access | Downloadable checkpoint | Downloadable under the Tencent Hy License |
| Operating limits | 32K-token context window | — |
| Sources | Model weights | Model weights |
| Inputs | Text | Text, Images |
| Weights | Open weights · MIT | Open weights · Tencent Hy License |
| Released | Nov 18, 2026 | Nov 18, 2026 |