Input knowledge
Upload source material — papers, notes, slides, docs — then pick format and run.
Classify a long document into selectable video topics.
Visual style Auto
Twelve templates, each built differently, with its own layout, font pair and way of moving: Whiteboard, Notebook, Chalkboard, Editorial, Magazine, Broadcast, Glass, Bento, Poster, Pop, Cinema and Terminal. Pick a colour on the strip.
Audience age + knowledge
Custom teaching prompt edit or reload
More options topics, publishing, slide images
We search the web (DuckDuckGo, Wikipedia, + optional Brave/Tavily/Serper) and pull short page extracts so slides get concrete examples — not just keywords.
Gemini studio voices narrate the whole video. Preview a card, then run — Gemma can still write the script locally.
Sample: “Hello! This is a quick sample of how this voice will sound…”
Tip: for books / long PDFs, use Plan topics then Run — one video or many, with web grounding.
Pipeline
Processing…Content review
Duplicates and voice ↔ highlight sync against the current script.
Generate slides to review them.
Record narration
Optional. Record a slide to replace Gemini narration on that slide only.
Run the pipeline first to generate a script, then record each slide in your own voice.
Slides
Rendered slide frames — expand to preview.
Slides
Rendered slide frames — expand to preview.
Publish details
Generate copy per platform when you need it — stored in this job so you can finish any tab later. Copy or download, then paste into the portal (or Post where credentials are set).
Caption, hashtags, and Instagram-sized slide images (carousel) or Reel.
Click Generate Instagram to create a post.
Twitter / X post with optional video or image.
Click Generate Twitter to create a post.
LinkedIn post for a professional audience.
Click Generate LinkedIn to create a post.
TikTok caption — generate when ready, then Copy / Download to paste in the app.
No API post yet — download the video from Outputs and paste this caption in TikTok.
Click Generate TikTok when you want a caption.
Full Substack newsletter (Markdown). Generate anytime — reuses this job’s narration & slides.
Paste into Substack editor (or upload the .md). No direct API publish.
Click Generate Substack to draft the newsletter.
Full Medium article (Markdown). Generate anytime from stored job context.
Paste into Medium (or upload the .md). No direct API publish.
Click Generate Medium to draft the article.
Outputs
Generated files.
Your pipeline and generated video will appear here.
Backend status
Cloud thumbnails via Gemini Nano Banana Pro. Local models not installed.
Run image test
Generate a test thumbnail with Nano Banana Pro (cloud). Local image models are not installed.
Preview
1280×720 Nano Banana Pro test output.
Generated image will appear here.
Session thumbnail
Regenerate a session's YouTube thumbnail (Nano Banana from .env).
Add user
| User | Access | Joined |
|---|
PUBLIC_BASE_URL for Graph API media.
Applies to scheduled papers and Run now. Studio uploads keep their own Visual style picker.
Everything needed to upload: title, description, tags, thumbnail, then Post (or Copy / Download).