Product Commercials with AI Video
Create cinematic product commercials entirely with AI. This guide walks you through the full Sequencer workflow: designing start and end frames, generating video with JSON prompts, and exporting in crisp 4K.
View JSON Prompt for This Commercial
How It Works
Product commercials follow a simple formula: a starting frame showing your product in a cinematic setting, a JSON prompt describing the motion and mood, and an ending frame, usually the brand logo. The AI generates the video between these two anchor images, creating a seamless cinematic transition.
This workflow uses Sequencer's node editor with the Nano Banana Pro model to generate the video, then exports at 4K for a polished, broadcast-ready result.
Start & End Frame Control
Anchor your commercial with a product hero shot and a brand logo. The AI generates fluid motion between them.
Structure your prompt as JSON to precisely control camera, lighting, motion, environment, and audio direction.
Generate at 1K for speed and iteration, then upscale to 4K on export for crisp, production-quality output.
Orbital pans, macro push-ins, and physics-aware movement. The model handles complex camera and subject motion.
Step 1: Create the Starting Frame
The starting frame is the first image your commercial opens on, typically a hero shot of your product in a cinematic environment. This is the visual anchor that sets the tone for the entire commercial.
Open the Image Editor in Sequencer and generate your product's opening shot. We recommend the Nano Banana Pro model for photorealistic product imagery with accurate lighting and materials.
Start at 1K resolution in the image editor. This keeps generation fast so you can iterate quickly on composition, lighting, and framing. We'll upscale to 4K later during export.
Write a prompt that describes your product in its environment. Be specific about the setting, lighting, and mood. For example:
A MacBook Air in Starlight aluminum floating among lush green fern fronds in a misty ancient forest, soft morning light diffusing through the mist, gentle highlights on the anodized aluminum curves, photorealistic product photography, shallow depth of field
Once you're happy with the composition, export the image. This will be your starting frame input for the video workflow node.
Step 2: Create the Ending Frame
The ending frame is what your commercial lands on, typically the product logo or brand mark. This gives the commercial a polished, professional close.
Repeat the same process: open the Image Editor, use Nano Banana Pro, and start at 1K resolution. This time, generate or place the brand logo in a complementary environment. The AI will generate video that transitions smoothly from your opening product shot to this ending frame.
For brand logos, try placing them in a visually connected environment. For example, if your product shot is set in a forest, your logo could appear over a tropical island or aerial landscape. The AI will create a seamless transition when the environments share visual logic.
Once generated, increase the resolution to 4K to sharpen the details, and export. This is your ending frame input node.
Step 3: Write the JSON Prompt
The JSON prompt is where you direct the commercial. It tells the AI exactly how to move the camera, light the scene, handle transitions, and design the audio.
Instead of a plain text prompt, product commercials benefit from a structured JSON prompt. This gives you precise control over every aspect of the shot. Here's the structure:
description
A high-level narrative of what happens in the commercial. Describe the mood, pacing, and overall story arc in 1-2 sentences.
style
The visual aesthetic. Reference real genres: "Apple Keynote aesthetic," "high-end product cinematography," "photorealistic 8K." This sets the rendering quality and look.
camera
Describe specific camera movements: "slow orbital pan," "smooth accelerated push-in," "seamless macro transition." Use real cinematography vocabulary.
lighting
Define the lighting for each scene or environment. Describe how it changes as the camera moves, e.g., "soft morning mist" outside → "clean studio lighting" inside.
environment
Map the journey: "Misty Ancient Forest → Clean Matte Black Internal Architecture → Tropical Apple Island." Use arrows to show transitions.
elements
List the physical objects that appear in the commercial. Be specific about materials, colors, and finishes, e.g., "Starlight Aluminum," "Matte black silicon with Glossy etched text."
sequences
Break the commercial into timestamped sequences. Each sequence has an "action" (what happens visually) and "audio" (what it sounds like). This gives the AI a clear shot-by-shot storyboard.
Full JSON Prompt Example (MacBook Air)
• Be cinematic, not technical. Describe what you'd see in a finished commercial, not how to render it.
• Use sensory language. Include sounds, textures, and physical behaviors: "premium dampened thud," "gentle highlights on aluminum curves."
• Map the journey. Use arrows in the environment field to guide the AI through scene transitions.
• Avoid generic descriptions. Instead of "a laptop on a table," write "the MacBook Air floats effortlessly among fern fronds, rotating slowly to catch the forest light."
Step 4: Export at 4K
Once your video looks good at 1K, increase the resolution to 4K on export to sharpen all the details. This is where Sequencer's upscale pipeline turns your fast iteration into a broadcast-ready commercial.
The workflow is designed so you can iterate quickly at lower resolution and only pay the cost of 4K when you're ready to finalize. This saves both time and credits while giving you maximum quality on the final output.
Workflow Summary
Starting Frame
Product hero shot (1K → 4K, Nano Banana Pro)
Ending Frame
Brand logo (1K → 4K, Nano Banana Pro)
Prompt Format
Structured JSON
Video Model
Any supported model
Export Resolution
4K (upscaled on export)
Clone our public Product Commercial workflow, drop in your product images, customize the JSON prompt, and start generating.