Audio-first creators
Podcast-to-video
Audio-first shows leave reach on the table on video-first platforms. direktor takes the episode script, narrates it, and generates a still per segment — a video version without a filming day.
input a podcast script or show notes → output narrated 1920×1080 MP4
Why direktor fits
- Your episodes are scripted or have detailed show notes to feed in as text.
- A narrated, illustrated cut is enough for YouTube and social distribution.
- You want each episode built the same way, from a CLI you can automate.
- You would rather generate segment visuals than license stock footage.
A typical run
- Save the episode script as a text file.
- Run direktor build episode.txt.
- Adjust the narration voice via BARK_MODEL if a different read suits the show.
- Post the MP4 alongside the audio release.
Honest caveat
direktor narrates from a script with a synthesised voice; it does not clone your existing podcast hosts or sync to their recorded audio.
Try it on your own text
The quickstart takes you from install to a finished MP4 in six steps.