Voice dictation: support multiple languages or separate language setting for speech-to-text

Problem

The microphone button in the VS Code extension only recognizes one language at a time for voice dictation. There is no way to configure the speech-to-text language independently — the only available setting (language) controls everything together: voice recognition, Claude's response language, and session titles.

This is a problem for users who work in multilingual contexts. For example, I speak Italian but produce English content (client emails, website copy). I need voice input to understand both languages, while Claude's response language should depend on context, not a global setting.

Proposed solution

One or more of the following:

• Multilingual speech-to-text that auto-detects the spoken language
• A separate setting for voice dictation language (e.g. dictation.language), independent from the response language
• Support for multiple languages in the dictation setting (e.g. dictation.languages=["it","en"])