AI Generation

Generate content using AI models.

Current Models

Images

Google Nano Banana - Image generation and editing
Supports aspect ratios: 1:1, 16:9, 9:16, 4:5, 3:2

Videos

Google Veo 3 Fast - Video generation
Duration: 5-8 seconds
Limited aspect ratio support

Captions

Gemini 2.5 Flash - Text generation for video captions

Generation Modes

Prompt-Based

Enter task/description
AI generates from text prompt
Uses brand/profile settings

Reference-Based

Upload reference image
AI creates similar content
Maintains style consistency

Manual Upload

Upload your own media
AI generates caption only
Applies brand voice

Generation Settings

Inherited from brand/profile:

system_prompt - AI instructions
caption_style - Writing style
brand_voice - Tone and personality
temperature - Creativity level (0-1)
hashtags - Auto-included tags
themes - Content topics

Content Types

Post Types

Feed - Standard posts
Story - 24-hour content
Reel - Short videos (Instagram)
Short - Short videos (YouTube)

Media Types

Image - Static images
Video - Motion content (5-8 sec)
Carousel - Multiple images (not implemented)

Platform Adaptations

Content automatically adjusted for:

Character limits
Hashtag formats
Platform-specific features
Aspect ratio requirements

Batch Generation

Create multiple content items:

Select multiple schedules
Click "Generate"
Processes in parallel
Track via Pusher updates

Image Editing

Edit generated images:

Available in review status
Enter edit prompt
Uses nano-banana model
Preserves original as backup

Generation Status

Pending - Waiting to start
Generating - In progress
Reviewing - Ready for approval
Failed - Error occurred
Scheduled - Approved and scheduled

API Endpoints

POST /api/content/generate
{
  "profileId": "uuid",
  "task": "prompt text",
  "generationMode": "prompt_based",
  "media": "image",
  "postType": "feed",
  "dimensions": "1:1"
}

POST /api/content/:id/regenerate
POST /api/content/:id/edit-image
{
  "prompt": "edit instructions"
}

Current Limitations

No prompt templates UI
No multi-language support
Basic tone control only
No custom model selection
Video limited to 5-8 seconds
Carousel not implemented
No A/B testing features

Real-time Updates

Generation progress via:

Pusher websockets
Status updates
Completion notifications

Next Steps

Overview Scheduling & Publishing