Modeling transcript export formats in a web app API

general
1

I have a programming question about modeling transcript export state in a web app.

The flow is:

  1. User submits a YouTube URL or video id.
  2. Server checks whether transcript or subtitle text is available.
  3. Client shows readable transcript text when available.
  4. User can copy text or export TXT.
  5. If timed data is available, SRT or VTT export can be offered separately.

The part I want to model cleanly is the response shape. TXT is readable text, while SRT/VTT is timed subtitle data, so I do not want the UI to imply that all formats are always available. https://aiyoutubetranscript.com/

One possible shape:

type TranscriptFormat = "txt" | "srt" | "vtt";

type TranscriptResponse = {
  source: "youtube";
  transcriptAvailable: boolean;
  language?: string;
  readableText?: string;
  availableExports: TranscriptFormat[];
  caveats: string[];
};

Would you keep export availability in one list like this, or split readable text and timed subtitle exports into separate objects so the UI cannot accidentally label TXT as a subtitle file?