fix: update Gemini latency reference to OpenRouter in release notes

Author: hannesrudolph

docs/update-notes/v3.19.6.mdx

@@ -4,7 +4,7 @@ This release introduces implicit caching for Gemini models to reduce latency, cl
 
 ## Provider Updates
 
-*   **Gemini Latency**: Replaced explicit caching with implicit caching for Gemini models to significantly reduce latency ([#4488](https://github.com/RooCodeInc/Roo-Code/pull/4488))
+*   **OpenRouter Latency**: Replaced explicit caching with implicit caching for Gemini models via OpenRouter to significantly reduce latency ([#4488](https://github.com/RooCodeInc/Roo-Code/pull/4488))
 
 ## QOL Improvements
 

docs/update-notes/v3.19.mdx

@@ -4,7 +4,7 @@ This release series focuses on streamlining the user interface with intelligent
 
 ## Provider Updates
 
-*   **Gemini Latency**: Replaced explicit caching with implicit caching for Gemini models to significantly reduce latency ([#4488](https://github.com/RooCodeInc/Roo-Code/pull/4488))
+*   **OpenRouter Latency**: Replaced explicit caching with implicit caching for Gemini models via OpenRouter to significantly reduce latency ([#4488](https://github.com/RooCodeInc/Roo-Code/pull/4488))
 
 *   **Bedrock**: Improved model information detection for custom Bedrock ARNs, ensuring better compatibility (thanks adamhill!)