FluidVoice v1.5.5

Biggest update yet - Test it and let me know what you think and if anything is broken xD

Implemented automatic media playback control (start/stop) during dictation and transcription [ Silicon Macs only for now ]
Introduced a download progress indicator for Voice models.
Added a customizable start chime for recording sessions, accessible via preferences.

Redesigned the entire AI Settings sidebar into dedicated Voice Engine and AI Enhancements tabs [Hope you like it!]
Added support for the upcoming Fluid-1 fully local model.
Integrated a model list retrieval button to automatically fetch available models from selected providers.
Expanded built-in provider support to include OpenAI, Google, Ollama, Cerebras, Groq, LM Studio, and OpenRouter.
Resolved various bugs and improved the overall stability of AI Settings.
Added direct links to obtain API keys for all supported providers within the settings interface.
Optimized the selection process for FluidVoice Voice Engines used in transcription and dictation.

Added the ability to define custom prompts within AI Settings for specialized tasks such as terminal commands or email templates.
Enabled on-the-fly prompt configuration through the dictation overlay.

Enhanced transcription engine can now process audio files exceeding 5 hours in length.
Improved performance to 160x real-time speed, allowing 5 hours of audio to be transcribed in approximately 2 minutes.

Removed the transcription history limit, increasing it from 1,000 entries to unlimited.
Verified compatibility with overlay-based applications such as Raycast.
Updated the UI to hide AI Enhancement menu items when the feature is disabled to reduce interface clutter.