Nano Banana + Veo Workflow#

Generate premium images and animate them into cinematic videos using Google's Gemini API.

Quick Start#

bash

# Generate just an image
uv run {baseDir}/scripts/generate.py \
  --image-prompt "iPhone floating against dark background, premium product photography" \
  --output-image hero.png

# Generate image + video
uv run {baseDir}/scripts/generate.py \
  --image-prompt "iPhone floating against dark background, premium product photography" \
  --video-prompt "iPhone gently rotating and floating, smooth seamless motion, cinematic lighting" \
  --output-image hero.png \
  --output-video hero.mp4 \
  --video-duration 4

Parameters#

ParameterRequiredDefaultDescription--image-prompt✅-Prompt for image generation--video-prompt❌image-promptPrompt for video animation--output-image✅-Output path for image (.png/.jpg)--output-video❌-Output path for video (.mp4)--resolution❌1KImage resolution: 1K, 2K, 4K--video-duration❌4Video seconds (4-8, Veo requirement)--video-resolution❌720pVideo quality: 720p, 1080p, 4k

Prerequisites#

• GEMINI_API_KEY environment variable set
• google-genai and requests Python packages

Installation#

bash

# Install dependencies
pip install google-genai requests

# Or with uv
uv pip install google-genai requests

Workflow Examples#

Hero Section Asset#

bash

uv run {baseDir}/scripts/generate.py \
  --image-prompt "Sleek iPhone 15 Pro showing Islamic prayer app interface, dark green and gold accents, floating against pure black background, dramatic studio lighting, premium product photography, 8K quality" \
  --video-prompt "iPhone floating and gently rotating in space, subtle purple glow, smooth seamless boomerang motion, premium cinematic lighting, dark background" \
  --output-image namazlock-hero.png \
  --output-video namazlock-hero.mp4 \
  --video-duration 4

Feature Showcase#

bash

uv run {baseDir}/scripts/generate.py \
  --image-prompt "Minimalist smartphone displaying prayer time interface, clean UI, dark mode with gold accents, studio lighting" \
  --output-image feature.png \
  --resolution 2K

Output#

• Images are saved in PNG format
• Videos are saved in MP4 format (H.264 encoded)
• Veo generation takes ~30-60 seconds (async polling handled automatically)

Notes#

• Veo 3.1 requires 4-8 second duration (enforced by API)
• Video generation is asynchronous — script handles polling automatically
• Use {baseDir} placeholder in paths — it resolves to the skill directory