
The Hook
“(No spoken dialogue — Silent video with ambient visual cues.)”
Product Reveal
“(No spoken dialogue — Silent video with ambient visual cues.)”
Feature Teaser
“(No spoken dialogue — Silent video with ambient visual cues.)”
Call to Action
“(No spoken dialogue — Silent video with ambient visual cues.)”
Synthesizable in Remotion · key components of the ai-agent-iterative-creation template.
Text appears character by character, simulating real-time typing of code or prompts within a terminal-like interface. This is often accompanied by a blinking cursor.
Use `interpolate` for `opacity` and `text-overflow: hidden` with `width` animation on a `<span>` or `<pre>` element. A `spring` animation can be used for the blinking cursor's opacity.
An image on screen undergoes a visible transformation or addition of elements directly within its frame, without a hard cut or transition to a new image.
Implement this with `Image` components where `src` is updated. For smooth transitions between image states, consider `crossFade` or `dissolve` effects, or layering elements with `opacity` and `scale` animations.
Small, colored dots (e.g., green for success) appear next to status messages, often fading in or pulsing briefly to indicate an action's completion.
Utilize `spring` or `interpolate` for `opacity` and `scale` on a small `div` element with `border-radius: 50%` to create the dot. A slight `pulse` effect can be achieved by animating `scale` between 1 and 1.1.
Purpose. Introduce the user's input and the AI's initial task.
Execution. Typing of a creative prompt in a code editor-like interface.
Purpose. Showcase the AI's ability to generate a base image from the prompt.
Execution. A blank canvas on the left side of the UI is filled with the first generated image.
Purpose. Demonstrate the agentic editing capabilities through sequential refinements.
Execution. Subsequent prompts are typed, and the image on the left updates in real-time with new elements and details.
Purpose. Present the final, complex image as a result of the iterative process.
Execution. The UI elements recede, leaving the fully refined image prominent, with a summary of edits.
Objective
Generate a detailed 3D model of an ancient alien artifact from a text description.How the recipe adapts
The left pane would display a 3D viewport instead of an image. Initial text prompts like 'Generate a weathered obsidian monolith with glowing runes' would appear via CodeTypewriterReveal. The 3D model would then appear and iteratively refine its textures, geometry, and lighting in the 3DInPlaceTransform primitive as subsequent prompts like 'Add a pulsating energy core' are typed. The final scene would show the completed 3D model slowly rotating, with a summary of the generation steps.Scene durations (4 scenes, 28s)
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