LLMClient implementation backed by Android's on-device Gemini Nano via the ML Kit GenAI Prompt API. More...
| Header: | #include <AndroidPlatformLLMClient> |
| Since: | 1.0 |
| Inherits: | LLMClient |
On Android targets this routes chat requests through a Java bridge (org.octomy.llm.AndroidGenAIBridge) that wraps the ML Kit GenAI Prompt API. The bridge dispatches token / completion / error events back through JNI native methods, which this class marshals to its owner thread and emits as standard LLMClient signals.
On non-Android targets the client compiles as a no-op stub: isAvailable() returns false and chat()/streamChat() emit an error so that the registry can still surface the provider in the UI without breaking desktop builds.
See also LLMClient and LLMManager.