How to apply caption presets (e.g. "Charged Up") via /v2/video/generate API?

Captions render with default styling.

Per the changelog from ~10 months ago, the text object was introduced for subtitle customization (presets, font size, position, etc.). I've tried passing various text configurations:

"text": { "preset_name": "Charged Up" }
"text": { "font_size": 80, "font_color": "#FFFFFF", "highlight_color": "#8B5CF6" }
The API accepts all of these without error, but the output always has default captions — no styling is applied.

Question: What is the correct schema for the text object to apply a caption preset like "Charged Up"? Is this feature currently functional via the API?

Thanks!