Added area for setting additional query parameters in API key input UI (e.g., ElevenLabs diarize parameter)

[gest-ctrl]

Hi there! I'll start by saying that I love this application and appreciate the sweat and tears that've gone into its development.

One idea to consider: several of the speech-to-text models accept additional parameters with the speech payload -- could there be an "Advanced" area or similar in the API key page where users can set these parameters?

For example, the ElevenLabs Speech-to-Text (“Scribe v1”) endpoint now exposes three parameters that can dramatically improve multi-speaker transcripts:

• diarize – boolean, default false. When enabled the service tags each word with a speaker_id, letting us attribute dialogue to individual speakers. Could be configured as an on/off toggle
• num_speakers -- integer, 1-32. The maximum amount of speakers talking in the uploaded file. Can help with predicting who speaks when. The maximum amount of speakers that can be predicted is 32.
• diarization_threshold – float 0.1 – 0.4. Higher values merge similar voices (fewer speakers); lower values split them (more speakers). This is honored only when diarize=true and num_speakers is unset; if omitted, ElevenLabs falls back to about 0.22. Could be conf

ElevenLabs API References - Speech-to-Text

I think a careful review of the various model providers could unearth a few useful parameters.

A potential downside to this change is the need to continually maintain a working set of parameters. But that's much more work than you'll already need to invest to maintain compatibility with the API syntax for each model provider.