AidGen SDK (Generative AI)
AidGen is an on-device inference SDK for generative AI. It supports local loading and inference of models such as LLM and VLM, and provides a C++ API for application integration.
Supported Capabilities
| Capability | Description |
|---|---|
| Model loading | Supports loading from a local model directory; no network required |
| Streaming output | Supports token-by-token streaming output, making typewriter effects easy to implement |
| Multimodality | Supports VLM models with text and image input |
| Hardware acceleration | Coordinated scheduling across CPU, GPU, and NPU |
