🎨 Studio 📁 Archive 🎬 Video Studio ⚗️ Experience 📱 App
Enter your xAI API key above
Saved Cards
Video Type
Camera + Grok Voice: Ken Burns / presets on your card + narration baked into the same file (supports 5–30s). Living Scene (AI): Imagine Video animates people & environment (API max 15s). Camera Move: silent local pan/zoom.
HARD RULE: never repeat text already in the image.
Video Style
Styles the Living Scene prompt and the energy of camera presets.
Image Source
Animation Preset
Modern motion presets — no extra text. Pair with Scene Motion for finished AI cards.
Duration
Camera / Camera+Voice: full length. Living Scene (AI): API allows 1–15s (30s clamps to 15).
Frame Rate
SOE Logo / Watermark
Voiceover Narration
Turning Grok Voice on switches Video Type to Camera + Grok Voice and muxes narration into the exported file.
Grok Voice
Use the dropdown or list. Tap ▶ to preview. Narration is muxed into Camera + Voice. If Auto-tags is on (Voiceover tab), Grok recommends speech tags (emphasis / persuade / etc.) before TTS.
Export File Type
WebM is native in the browser. MP4 re-encodes when the browser supports it (Safari often does); otherwise falls back to WebM.
Select a card — Living Scene (AI) moves people & environment for the principle. Never repeats text.
Animation Preview
🎞
Select a saved card to animate
Load from Archive
Principle Short Story
Pick a principle to build an 8s short-story arc (setup → turn → payoff).
Story = 3-beat principle film. Clip = single moment (legacy). Continuity Bible always injects when items are pinned/favorited.
Continuity Bible
ADD / EDIT REFERENCE
Video Style
Duration
xAI Imagine Video supports 1–15 seconds per clip.
Aspect Ratio
Attachments (optional)
An attached image is sent as the starting frame (image-to-video). Continuity images help lock faces/places.
Video Prompt ✏ Editable
Download File Type
xAI returns MP4. Choosing WebM re-wraps/re-encodes in-browser when possible.
Voiceover (after video)
Uses the same voice as Voiceover tab (including your xAI clone). Downloads a separate MP3; video stays silent unless you mix in an editor.
⚠ AI video is billed by the second (Imagine Video). Pick duration above (up to 15s). Polling may take 1–3 minutes.
Load a card or type a prompt to begin
Generated Video
🤖
AI-generated video will appear here
Text to Speak
Speech Tags (Grok Voice)
xAI TTS supports inline tags like [pause] [sigh] and wrapping tags like <emphasis> <slow> <build-intensity>. Grok recommends tags for your delivery intent, then we send the tagged script to TTS.
Delivery intent
Tagged script (editable · sent to TTS)
Tags are applied sparingly for natural SOE narration — not every sentence.
Grok Voice (full roster)
Tap ▶ to hear a sample before you commit.
Speed
Archive Card (optional)
Enter text and click Generate
Voiceover Preview
🎙
Generated audio will appear here