Feature
- long press volume key to show the volume adjustment panel, when in page turn mode
- support more web pages to be able to be translated by paragraphs
- add input url action for gesture action
- change default gemini llm model to gemini-2.5-flash
- zoom feature is supported in fullscreen mode (not perfect though)
- change reasoning level to none for openai API usage
Issue fixes
- context menu dialog fragment is not attached sometimes, and leads to crash