Table 4. Features of Inference Engine and OpenAI API for LLM Development Environment in Unity.

	Features
Inference Engine	- Running locally (Hardware dependency)- LLM file must included in the project- LLM quantization may needed- Tokenizer required- Free (using open-source LLMs)- Using LLM in ONNX format
OpenAI API	- Using RESTful API to access OpenAI server- Ease of use- Accessing various OpenAI models and features (including RealTime API)- Come with costs- Unofficial Unity package