Run and deploy your AI Studio app

This will automatically set AI Models:

• using Gemini (specifically the gemini-2.5-flash-image model) for generating and editing images.
• using Veo (specifically the veo-3.1-fast-generate-preview model) for generating videos ( NOT Free).

This contains everything you need to run your app locally.

View the app in Github Pages (requires Gemini API Key) : https://ai-studio-apps.github.io/GenerativeNodeFlow/

View the app in AI Studio: https://aistudio.google.com/apps/drive/1qLISjOuHyN53q77FUVNPLmrLlZ8auOHr

Run Locally

Prerequisites: Node.js

1. Install dependencies: npm install
2. Set the GEMINI_API_KEY in .env.local to your Gemini API key
3. Run the app: npm run dev

AI Model Configuration

This app allows you to configure different Gemini models for specific tasks (Text, Image, Video) via the Settings panel.

• Independent Settings: You can define a unique API Key, Model Name (e.g., gemini-2.5-flash-image, veo-3.1-fast-generate-preview), and Base URL for each modality using the tabbed interface.
• Persistence: Your configurations are saved automatically to your browser's Local Storage, preserving your setup between sessions.
• Dynamic Execution: The application dynamically selects the correct credentials and model version based on the active node type (Text Generator, Image Editor, or Video Generator).

Tip

Open the app settings to specify the API Key, Model, and URL (local or cloud)
These settings are saved to your localstorage.

• Model for text.
• Model for generating and editing images.
• Model for generating videos.

Using Google Gemini

You only need one single API key (your Google GenAI / Gemini API key).

Unified Access: Both the Gemini and Veo models are accessed through the same Google GenAI SDK (@google/genai).

One Key: The specific API Key you provide in your .env file (GEMINI_API_KEY) authenticates your project with Google's services, granting you access to all available models (Gemini, Veo, Imagen, etc.) that your account is authorized to use.

Caution

Note on Billing: While you use the same key, Veo (Video generation) is a premium feature. To use Veo models, the Google Cloud Project associated with your API key must usually have billing enabled (meaning it requires a paid tier), whereas some Gemini models have a free tier

Test local build of metadata.json

This project builds in CI using npm. If you prefer pnpm for local development you may still use it, but CI runs npm for consistency across runners. You can enable pnpm locally using Corepack (optional) or install it globally.

Quick commands (Windows / macOS / Linux)

Use npm locally to match CI. The following commands install dependencies, build the site, and generate dist/metadata.json.

PowerShell (Windows):

# Install dependencies and build
npm install
npm run build

# Generate metadata.json into dist and view it
node scripts/generate-metadata.cjs
Get-Content .\dist\metadata.json -Raw

Bash (macOS / Linux):

# Install dependencies and build
npm install
npm run build

# Generate metadata.json into dist and view it
node scripts/generate-metadata.cjs
cat dist/metadata.json

Note

CI uses npm for installs and builds. For deterministic CI installs, add and commit package-lock.json so the workflow can use npm ci.