
AI avatars just gained full-body movement. The result looks surprisingly human. For years, AI avatars were limited to talking heads - locked in one frame, one angle. Now they can move through 3D sp…
The Hook
“Before I could only talk, now I can act.”
Problem Agitation
“You can prompt my outfit, my background, or my actions. Just like in real life.”
Product Reveal
“It's that easy. Select which avatar to create looks for. Blue. Prompt the. Or. A modern, corporate training environment: an empty, well-list seminar room with rows of minimalist... through her binoculars in a national park at sunset. from cinematic lighting and natural motion.”
Feature Teaser
“Watch it come to life.”
Call to Action
“Finally, customize avatars and make them act. Try it today in Synthesia.”
Synthesizable in Remotion · key components of the dynamic-ai-avatar-launch template.
Text appears with a quick, impactful fade-in and slight scale-up, often accompanied by a sound effect. It can be used for single words or short phrases.
Use `spring` for `opacity` and `scale` properties. Interpolate `from: 0` to `to: 1` for both. Apply a short duration and a high damping ratio for a snappy effect. Coordinate with audio cues.
Rapid succession of short video clips or images, showcasing diverse scenarios or features. Each clip is typically 1-2 seconds long, creating a high-energy visual flow.
Implement a sequence of `Video` or `Img` components within a `Sequence`. Use `AbsoluteFill` for each scene. Control timing precisely with `durationInFrames` and `from` props. Ensure smooth, hard cuts between scenes.
A mouse cursor animates across the screen, clicking on specific UI elements, highlighting them, and triggering subsequent interface changes or content reveals.
Utilize `interpolate` for `x` and `y` coordinates of a custom cursor SVG or image. Use `spring` for `scale` on click events to simulate button presses. Overlay this animation on top of screen recordings or mockups.
Purpose. Immediately establishes the core problem and the product's transformative solution, hooking the viewer with a clear value proposition.
Execution. Presenter states 'Before I could only talk, now I can act' with corresponding text overlays and a visual transformation from static to dynamic.
Purpose. Demonstrates the product's capabilities through diverse, engaging scenarios, proving the 'full-body movement' claim.
Execution. A rapid montage of the AI avatar performing various actions in different environments (supermarket, construction, space).
Purpose. Educates the viewer on the ease of use and customization options, building confidence in the product's accessibility.
Execution. Screen recording of the Synthesia interface, showing avatar selection, text prompting, and media generation with cursor interactions.
Purpose. Reinforces the product's impact and guides the viewer to the next step, solidifying the brand message.
Execution. A grid of dynamic avatar videos, followed by the presenter's final call to action and the Synthesia logo reveal.
Objective
Launch a new data visualization platform called 'QuantumFlow' that transforms static spreadsheets into interactive 3D models.How the recipe adapts
The 'Before' scene would show a static spreadsheet on screen, then transition to a 'Now' scene where data points burst into a dynamic, flowing 3D visualization. Instead of avatar actions, the montage would feature different data sets being visualized in unique ways (e.g., financial trends as a city skyline, scientific data as a molecular structure). The UI walkthrough would demonstrate dragging and dropping data, selecting visualization types, and interacting with the 3D models. The final call to action would be 'Visualize your data today with QuantumFlow'.Scene durations (5 scenes, 54s)
Key Remotion components
To ensure that artificial general intelligence benefits all of humanity

Gopuff introduces Go, an AI shopping assistant built with SpaceXAI: say what you need and the order is placed.

Iliad (Troy) trailer made by Grok Imagine 1.5, which was just released