Create inference-provider.md

0_app/4_api/inference-provider.md

@@ -0,0 +1,24 @@
+# How to add LM Studio as an LLM provider
+
+While LM Studio has an OpenAI-compatible API, it also provides SDKs for Python and TypeScript/JavaScript, offering a richer set of features tailored specifically for local AI development.
+
+## Why use LM Studio’s SDK over OpenAI’s SDK?
+
+LM Studio’s SDKs are designed to be clean and developer-friendly for Python and TypeScript/JavaScript users.  
+
+They include a number of local AI-focused features not found in OpenAI's SDK, such as:
+
+- **Model Management**: Load, unload, and manage AI models directly from memory  
+- **Model Information**: Get details like context length and model size  
+- **Reliable Request Handling**: Easily abort ongoing prediction requests  
+- **Advanced Configuration**: Customize model load parameters and use Presets (e.g., GPU offloading, context length)  
+- **Speculative Decoding**: Use advanced inference techniques for improved performance
+
+Using LM Studio’s SDKs also means your application can automatically connect to LM Studio without needing to manually start a server or configure ports.
+
+## Getting Started
+
+Adding support for LM Studio as an inference provider is easy:
+
+- [TypeScript/JavaScript Setup →](#)
+- [Python Setup →](#)