In the darkroom. This engine is still being developed and tested; details can change.
WS-024 · FR 24 · Engine Room
Explainer Engine
A read-aloud text and your images in, an animated explainer on the voice out.
A · Free · keyword EXPLAIN
The preset and the shotlist
The preset this engine reads and the shotlist template it fills. Enough to run one shoot by hand and see if it fits your brand.
Fig. 1 · The board
A Input slot
a read-aloud text (20-45 s) + images and clips in assets/library/
B Engine
- 01explainer
- 02voice-sound
- 03humanizer
route local · fallback gemini
C Contact sheet
explainer.md (timeline), voice.mp3, final.mp4 in 9:16 with sound effects and optional music
Data sheet
- Input
- a read-aloud text (20-45 s) + images and clips in assets/library/
- Output
- explainer.md (timeline), voice.mp3, final.mp4 in 9:16 with sound effects and optional music
- Models
- elevenlabs
- Route
- local (fallback gemini)
- Command
- /explainer
- Cost per run
- Local and free with the mock voice; a real voice-over costs ElevenLabs characters (about 75 words per 30 s)
- Maturity
- Working: used in production here
Fig. 2 · Sample output
3 frames · real output from this engine
All images on this page are generated with this engine.
What it does
It animates an explainer on the voice: every word lands when it is heard, one big word per phrase, a dark board that
fills chapter by chapter, and a wall of cards that gives way to one hero. The images and clips come from the asset
library; colours and fonts from the explainer block in tokens.json.
What you see in the post
A wall of fashion frames, then “This red dress was never shot” over one frame, then the board with the casting card and the four keyframes landing on the voice.
How it runs
- The read-aloud text (delta hook, result first, word budget) through
humanizer; approval of the text. - Voice first:
core/voice/tts.py(mock free, ElevenLabs after the character count is approved). explainer.md: per beat word stream, board or wall to hero, with cards fromassets/library/.core/explainer/explainer.py: Chrome draws every frame locally, then the mix with sound effects and music.- QC frames per beat; approval of the picture.
What you get (the giveaway)
- free: the read-aloud template (delta hook, result first, one idea per beat, word budget per length)
(
../weadapt-studios-products/product/studios/video-edit-studio/free/) - studio: the Video Edit Studio for Claude Code and Codex
Honest limits
Runs on our own explainers every week, with a real voice. It explains with what is in the library and makes no new images. In the built studio it renders in the buyer’s tokens; a buyer-style run with a real voice is still to do.
