The ailia SDK is an inference engine that supports cross-platform operation. ONNX can perform GPU inference using Metal and Vulkan. It offers a model library with over 400 types and supports advanced pre-processing and post-processing. Bindings are provided for Python, C++, C#, and Flutter.
Full Documentation Site (docs.ailia.ai)
- ailia SDK - ONNX inference API (Python, C++, Unity, Flutter, JNI)
- ailia Tokenizer - Tokenization for NLP text encoding and decoding
- ailia Speech - Speech recognition (Speech2Text)
- ailia Voice - Text-to-speech synthesis
- ailia LLM - Local LLM inference with GGUF support
- ailia Tracker - Object movement tracking
- ailia TFLite - TFLite inference