Fixed a problem with new text-shadow default value
0.6.0
Default voice model is now gpt-4o-mini. Higher quality and cheaper! Slightly increased default text font size. Text background now defaults to text shadow, instead of a box background.
0.5.1
fixed crash when trying to use obscure characters
0.5.0
- Slightly improved video rendering speed - Slightly reduced video file size - Improved character detection in user prompts (fuzzy string matching)
0.4.4
Added support for aac audio codec.
(And any other coded supported by ffmpeg, but without type hints)
0.4.3
Fixes minor background music bug in previous version where the URL was not always returned properly