google-gemini-image-video-generation

id	ailey-media-gemini
name	Google Gemini Image & Video Generation
description	Comprehensive Google Gemini integration for AI-powered image and video generation using Imagen 3 and Veo 2. Create, edit, upscale images and generate videos from text prompts.
keywords	["gemini","google","imagen","veo","image-generation","video-generation","ai","media"]

Google Gemini Image & Video Generation Skill

Comprehensive Google Gemini integration for AI-powered image and video generation using Imagen 3 and Veo 2. Create, edit, upscale images and generate videos from text prompts.

Overview

Harness Google's latest generative AI models for visual content creation:

• Imagen 3: State-of-the-art text-to-image generation with photorealistic quality
• Veo 2: Advanced text-to-video generation with cinematic quality
• Image Editing: Inpainting, outpainting, and style transfer
• Upscaling: 4x resolution enhancement
• Multiple Account Tiers: Free, Pay-as-you-go, and Enterprise (Vertex AI)
• Flexible Output: PNG, JPEG, MP4, WebM formats
• Rate Limit Management: Automatic tier-based throttling
• Metadata Tracking: Generation history and parameters
• Batch Operations: Process multiple prompts efficiently

Account Tiers

Free Tier

• Cost: $0 (with usage limits)
• Authentication: API key from Google AI Studio
• Rate Limits: 15 requests/minute, 1,500 requests/day
• Models: Imagen 3 (basic), Veo 2 (limited)
• Features:
  • Text-to-image generation
  • Basic video generation (up to 8 seconds)
  • Standard resolution (up to 1024x1024)
  • Basic editing capabilities
• Best For: Hobbyists, testing, personal projects
• Signup: https://aistudio.google.com/

Pay-as-you-go Tier

• Cost: Usage-based pricing
  • Imagen 3: $0.020 per image (standard), $0.040 (high quality)
  • Veo 2: $0.30 per video (8 sec), $0.60 (16 sec)
  • Upscaling: $0.010 per image
• Authentication: API key from Google AI Studio
• Rate Limits: 60 requests/minute, 10,000 requests/day
• Models: Full Imagen 3, Veo 2 access
• Features:
  • All Free tier features
  • High-quality image generation
  • Extended video generation (up to 16 seconds)
  • 4K resolution support
  • Advanced editing (inpainting, outpainting)
  • Batch operations
  • Priority processing
• Best For: Content creators, businesses, production use
• Signup: https://aistudio.google.com/ → Enable billing

Enterprise Tier (Vertex AI)

• Cost: Custom pricing, volume discounts
• Authentication: Service Account + OAuth
• Rate Limits: 300 requests/minute, 100,000+ requests/day
• Models: All models + early access to new features
• Features:
  • All Pay-as-you-go features
  • 8K resolution support
  • Longer videos (up to 60 seconds)
  • Custom model fine-tuning
  • Advanced safety filters
  • SLA guarantees
  • Dedicated support
  • VPC integration
  • Audit logging
• Best For: Enterprises, high-volume production
• Signup: https://cloud.google.com/vertex-ai → Contact sales

Feature Matrix

Feature	Free	Pay-as-you-go	Enterprise
Imagen 3 (Text-to-Image)	✅ Basic	✅ Full	✅ Full
Veo 2 (Text-to-Video)	⚠️ Limited	✅ Full	✅ Extended
Max Image Resolution	1024x1024	2048x2048	8192x8192
Max Video Length	8 seconds	16 seconds	60 seconds
Image Editing	✅ Basic	✅ Advanced	✅ Advanced
Upscaling (4x)	❌	✅	✅
Batch Operations	⚠️ Limited	✅	✅
Rate Limit	15/min	60/min	300/min
Daily Limit	1,500	10,000	100,000+
Priority Processing	❌	✅	✅
Custom Fine-tuning	❌	❌	✅
SLA Guarantees	❌	❌	✅
Support	Community	Email	Dedicated

Getting Started

Step 1: Create Google Account

For Free/Pay-as-you-go:

1. Visit https://aistudio.google.com/
2. Sign in with Google account
3. Accept terms of service
4. No credit card required for Free tier

For Enterprise (Vertex AI):

1. Visit https://cloud.google.com/vertex-ai
2. Create or select GCP project
3. Enable Vertex AI API
4. Set up billing account
5. Contact sales for enterprise features

Step 2: Get API Key

For Free/Pay-as-you-go:

1. Go to https://aistudio.google.com/app/apikey
2. Click "Create API key"
3. Select project (or create new)
4. Copy API key
5. Important: Keep key secure, never commit to version control

For Enterprise (Vertex AI):

1. Create Service Account in GCP Console
2. Grant necessary permissions:
   • aiplatform.user
   • storage.admin (for outputs)
3. Download JSON key file
4. Set environment variable: GOOGLE_APPLICATION_CREDENTIALS

Step 3: Enable Billing (Optional - for Pay-as-you-go)

1. Google AI Studio → Settings → Billing
2. Link Cloud Billing account
3. Set budget alerts (recommended)
4. Monitor usage at https://console.cloud.google.com/billing

Step 4: Install Skill

cd .github/skills/ailey-media-gemini
node install.cjs

Step 5: Configure API Key

Create .env file:

# For Free/Pay-as-you-go
GEMINI_API_KEY=AIzaSy...your_api_key_here
GEMINI_ACCOUNT_TYPE=free  # or 'paid'

# For Enterprise (Vertex AI)
GEMINI_PROJECT_ID=your-gcp-project
GEMINI_LOCATION=us-central1
GEMINI_SERVICE_ACCOUNT_KEY=/path/to/service-account-key.json
GEMINI_ACCOUNT_TYPE=enterprise

Configure in AI-ley (.github/aicc/aicc.yaml):

skills:
  gemini:
    type: media
    path: .github/skills/ailey-media-gemini
    config:
      apiKey: ${GEMINI_API_KEY}
      accountType: ${GEMINI_ACCOUNT_TYPE}
      outputDir: ./output/gemini
      defaultModel: imagen-3.0-generate-001

Usage Examples

Detect Account Tier

npm run detect

Generate Image (Text-to-Image)

npm run generate -- \
  --prompt "A serene mountain landscape at sunset" \
  --aspect-ratio "16:9" \
  --output "mountain.png"

Generate Video (Text-to-Video)

npm run video -- \
  --prompt "A drone flying over a futuristic city" \
  --duration 8 \
  --output "city-flyover.mp4"

Edit Image (Inpainting)

npm run edit -- \
  --image "photo.jpg" \
  --mask "mask.png" \
  --prompt "Replace with a sunset sky" \
  --output "edited.png"

Upscale Image

npm run upscale -- \
  --image "low-res.jpg" \
  --factor 4 \
  --output "high-res.png"

View Generation History

npm run history -- --limit 20

API Reference

GeminiClient Class

import { GeminiClient } from './.github/skills/ailey-media-gemini/src';

const client = new GeminiClient({
  apiKey: process.env.GEMINI_API_KEY,
  accountType: 'free' // or 'paid', 'enterprise'
});

Account & Configuration

detectAccountTier(): Promise<AccountTier>

Detect account tier and capabilities.

Returns:

{
  tier: 'Free' | 'Pay-as-you-go' | 'Enterprise',
  rateLimit: number,          // requests per minute
  dailyLimit: number,         // requests per day
  maxImageResolution: string, // e.g., "1024x1024"
  maxVideoLength: number,     // seconds
  features: string[]
}

Image Generation (Imagen 3)

generateImage(options): Promise<GeneratedImage>

Generate image from text prompt.

Parameters:

{
  prompt: string,
  negativePrompt?: string,
  aspectRatio?: '1:1' | '16:9' | '9:16' | '4:3' | '3:4',
  model?: 'imagen-3.0-generate-001' | 'imagen-3.0-fast-001',
  numberOfImages?: number, // 1-4 (tier-dependent)
  seed?: number,           // for reproducibility
  outputFormat?: 'png' | 'jpeg'
}

Returns:

{
  images: Array<{
    data: Buffer,
    mimeType: string,
    metadata: {
      model: string,
      prompt: string,
      seed: number,
      timestamp: string
    }
  }>,
  promptId: string
}

editImage(options): Promise<GeneratedImage>

Edit existing image (inpainting/outpainting).

Parameters:

{
  image: Buffer | string,      // image data or path
  mask?: Buffer | string,      // mask for inpainting
  prompt: string,
  mode: 'inpaint' | 'outpaint' | 'style-transfer',
  strength?: number,           // 0-1, edit intensity
  outputFormat?: 'png' | 'jpeg'
}

upscaleImage(options): Promise<GeneratedImage>

Upscale image resolution (4x).

Parameters:

{
  image: Buffer | string,
  factor: 2 | 4,
  outputFormat?: 'png' | 'jpeg'
}

Video Generation (Veo 2)

generateVideo(options): Promise<GeneratedVideo>

Generate video from text prompt.

Parameters:

{
  prompt: string,
  duration: number,            // seconds (tier-dependent max)
  aspectRatio?: '16:9' | '9:16' | '1:1',
  model?: 'veo-2.0-generate-001',
  fps?: number,                // 24, 30, 60
  outputFormat?: 'mp4' | 'webm'
}

Returns:

{
  video: Buffer,
  mimeType: string,
  duration: number,
  resolution: string,
  fps: number,
  metadata: {
    model: string,
    prompt: string,
    timestamp: string
  }
}

History & Management

getGenerationHistory(limit?: number): Promise<GenerationRecord[]>

Retrieve generation history.

saveOutput(data: Buffer, filename: string): Promise<string>

Save generated content to file.

Workflows

Workflow 1: Create Marketing Image

# 1. Generate hero image
npm run generate -- \
  --prompt "Professional product photo of smartwatch on marble surface" \
  --aspect-ratio "16:9" \
  --output "hero.png"

# 2. Upscale for print
npm run upscale -- \
  --image "hero.png" \
  --factor 4 \
  --output "hero-4k.png"

# 3. Create variations
npm run generate -- \
  --prompt "Professional product photo of smartwatch on marble surface" \
  --aspect-ratio "16:9" \
  --seed 12345 \
  --output "hero-v2.png"

Workflow 2: Generate Social Media Content

# Instagram post (1:1)
npm run generate -- \
  --prompt "Vibrant food photography, overhead shot" \
  --aspect-ratio "1:1" \
  --output "instagram.png"

# Instagram story (9:16)
npm run generate -- \
  --prompt "Vibrant food photography, overhead shot" \
  --aspect-ratio "9:16" \
  --output "story.png"

# YouTube thumbnail (16:9)
npm run generate -- \
  --prompt "Vibrant food photography, overhead shot" \
  --aspect-ratio "16:9" \
  --output "youtube.png"

Workflow 3: Create Product Demo Video

# 1. Generate product showcase video
npm run video -- \
  --prompt "Smooth product rotation of smartphone, studio lighting" \
  --duration 8 \
  --aspect-ratio "16:9" \
  --output "product-demo.mp4"

# 2. Generate B-roll footage
npm run video -- \
  --prompt "Close-up details of smartphone features and interface" \
  --duration 8 \
  --output "b-roll.mp4"

Workflow 4: Image Editing Pipeline

# 1. Generate base image
npm run generate -- \
  --prompt "Modern office interior" \
  --output "base.png"

# 2. Edit specific area (inpainting)
npm run edit -- \
  --image "base.png" \
  --mask "window-mask.png" \
  --prompt "Large windows with city view" \
  --mode "inpaint" \
  --output "edited.png"

# 3. Expand canvas (outpainting)
npm run edit -- \
  --image "edited.png" \
  --mode "outpaint" \
  --prompt "Continue the office space" \
  --output "expanded.png"

Workflow 5: Batch Content Creation

import { GeminiClient } from './.github/skills/ailey-media-gemini/src';

const client = new GeminiClient({
  apiKey: process.env.GEMINI_API_KEY!
});

const prompts = [
  "Mountain landscape at sunrise",
  "Ocean waves at sunset",
  "Forest path in autumn"
];

for (const prompt of prompts) {
  const result = await client.generateImage({
    prompt,
    aspectRatio: '16:9',
    numberOfImages: 2
  });
  
  result.images.forEach((img, i) => {
    await client.saveOutput(
      img.data,
      `${prompt.replace(/\s+/g, '-')}-${i + 1}.png`
    );
  });
}

Models

Imagen 3 (Image Generation)

• Model ID: imagen-3.0-generate-001
• Fast Model: imagen-3.0-fast-001 (lower quality, faster)
• Max Resolution: Tier-dependent (1024x1024 to 8192x8192)
• Aspect Ratios: 1:1, 16:9, 9:16, 4:3, 3:4
• Features: Photorealistic, artistic styles, text rendering

Veo 2 (Video Generation)

• Model ID: veo-2.0-generate-001
• Max Duration: Tier-dependent (8-60 seconds)
• Resolution: Up to 4K (3840x2160)
• Frame Rates: 24, 30, 60 FPS
• Features: Cinematic quality, camera movements, temporal consistency

TypeScript Integration

import { GeminiClient } from './.github/skills/ailey-media-gemini/src';
import fs from 'fs';

const client = new GeminiClient({
  apiKey: process.env.GEMINI_API_KEY!,
  accountType: 'paid'
});

// Generate image
const imageResult = await client.generateImage({
  prompt: "A futuristic cityscape at night",
  aspectRatio: '16:9',
  numberOfImages: 4
});

// Save all variations
for (let i = 0; i < imageResult.images.length; i++) {
  const path = await client.saveOutput(
    imageResult.images[i].data,
    `cityscape-${i + 1}.png`
  );
  console.log(`Saved: ${path}`);
}

// Generate video
const videoResult = await client.generateVideo({
  prompt: "Drone flying through the futuristic city",
  duration: 16,
  aspectRatio: '16:9',
  fps: 60
});

await client.saveOutput(videoResult.video, 'city-flyover.mp4');

// Upscale favorite image
const upscaled = await client.upscaleImage({
  image: 'cityscape-1.png',
  factor: 4
});

await client.saveOutput(upscaled.images[0].data, 'cityscape-4k.png');

Troubleshooting

Issue	Solution
"Invalid API key"	Verify API key at https://aistudio.google.com/app/apikey
"Quota exceeded"	Check usage limits, upgrade to Pay-as-you-go, or wait for reset
"Rate limit exceeded"	Implement exponential backoff, reduce request frequency
"Image generation failed"	Refine prompt, avoid prohibited content, check safety filters
"Video too long"	Reduce duration to tier limit (Free: 8s, Paid: 16s, Enterprise: 60s)
"Resolution not supported"	Check tier max resolution, downscale request
"Service account error"	Verify credentials path, check IAM permissions
"Billing not enabled"	Enable billing at https://console.cloud.google.com/billing

Safety & Content Policy

Google Gemini enforces content safety filters:

• Prohibited: Violence, hate speech, adult content, illegal activities
• Restricted: Political figures, copyrighted characters, deepfakes
• Guidelines: https://ai.google.dev/gemini-api/docs/safety-settings

Pricing Reference

Imagen 3 (per image)

• Standard quality: $0.020
• High quality: $0.040

Veo 2 (per video)

• 8 seconds: $0.30
• 16 seconds: $0.60
• 60 seconds (Enterprise): $2.40

Upscaling

• 4x upscale: $0.010 per image

Note: Prices current as of February 2026. Check https://ai.google.dev/pricing for latest.

Resources

• Google AI Studio: https://aistudio.google.com/
• API Documentation: https://ai.google.dev/gemini-api/docs
• Vertex AI: https://cloud.google.com/vertex-ai
• Pricing: https://ai.google.dev/pricing
• Safety Guidelines: https://ai.google.dev/gemini-api/docs/safety-settings

---

version: 1.0.0 updated: 2026-02-01 reviewed: 2026-02-01 score: 4.5