Pixeldojo

PixelDojo Skill

Generate AI images and videos using the PixelDojo API platform.

Setup

1. Get API Key: Sign up at https://pixeldojo.ai/api-platform
2. Buy Credits: Purchase credits at https://pixeldojo.ai/api-platform/buy-credits
   • $5 = ~800 images with P-Image model
3. Set Environment Variable:

   export PIXELDOJO_API_KEY=your_api_key_here
   

   (Or copy .env.example to .env and fill in your key)

• API Documentation: https://pixeldojo.ai/api-platform
• Base URL: https://pixeldojo.ai/api/v1

Available Models

Image Models

• flux-1.1-pro - Latest pro model with enhanced quality (recommended)
• flux-1.1-pro-ultra - Highest quality Flux model with raw mode
• flux-dev - Development model with configurable steps and LoRA support
• flux-krea-dev - Photorealistic, avoids oversaturated AI look
• flux-kontext-pro - Advanced model with editing capabilities
• flux-kontext-max - Premium model with maximum performance

Video Models

• wan-2.6-flash - Fast video generation
• wan-2.2 - Higher quality video
• veo-3.1 - Google's Veo 3.1 with native audio
• kling-2.5-turbo-pro - Kling turbo
• kling-pro - Kling professional
• minimax - Minimax video

Core Operations

Generate Image

# Basic image generation
bash ~/.openclaw/skills/pixeldojo/generate.sh image "a serene mountain landscape at sunset" flux-2

# With options
bash ~/.openclaw/skills/pixeldojo/generate.sh image "cyberpunk city" flux-2 --aspect-ratio 16:9 --output ~/Desktop/cyberpunk.png

Generate Video

# Text-to-video
bash ~/.openclaw/skills/pixeldojo/generate.sh video "ocean waves crashing on rocks" wan-2.6-flash --duration 5

# Image-to-video
bash ~/.openclaw/skills/pixeldojo/generate.sh video "make it cinematic" wan-2.6-flash --image-url https://example.com/image.png --duration 5

Check Job Status

bash ~/.openclaw/skills/pixeldojo/status.sh job_abc123

List Available Models

bash ~/.openclaw/skills/pixeldojo/models.sh

API Reference

Submit Job

• Endpoint: POST /api/v1/models/{model}/run
• Headers: Authorization: Bearer {API_KEY}, Content-Type: application/json
• Body (Image):

  {
    "prompt": "description of image",
    "aspect_ratio": "16:9"
  }
  

• Body (Video):

  {
    "prompt": "description of video",
    "image_url": "https://...",  // optional, for image-to-video
    "duration": 5,                // seconds
    "aspect_ratio": "16:9"
  }
  

• Response:

  {
    "jobId": "job_abc123",
    "status": "pending",
    "statusUrl": "https://pixeldojo.ai/api/v1/jobs/job_abc123"
  }
  

Check Status

• Endpoint: GET /api/v1/jobs/{job_id}
• Headers: Authorization: Bearer {API_KEY}
• Response (completed):

  {
    "jobId": "job_abc123",
    "status": "completed",
    "output": {
      "image": "https://temp.pixeldojo.ai/...png",
      "video": "https://temp.pixeldojo.ai/...mp4"
    }
  }
  

Download Result

Results are automatically downloaded to ~/Pictures/AI Generated/ with timestamped filenames for easy access.

Workflow

1. Submit: Call generate endpoint, get job ID
2. Poll: Check status URL every 2-5 seconds
3. Download: Once status === "completed", download from output URL
4. Save: Store in workspace with descriptive filename

Error Handling

• Rate limit: 10 requests/minute (contact support for higher)
• Credits: Check dashboard for credit balance
• Timeouts: Video generation can take 30-120 seconds

Example Usage

# Generate a cyberpunk portrait
bash ~/.openclaw/skills/pixeldojo/generate.sh image "cyberpunk samurai with neon lights, detailed, 8k" flux-2 --output ~/Desktop/samurai.png

# Animate a photo
bash ~/.openclaw/skills/pixeldojo/generate.sh video "slow motion drift, cinematic" wan-2.6-flash --image-url https://mycdn.com/photo.jpg --duration 5

Output Location

All generated files are saved to your Pictures folder for easy access:

• ~/Pictures/AI Generated/images/
• ~/Pictures/AI Generated/videos/

With format: {timestamp}_{prompt_snippet}.{ext}

Use --output <path> to save to a custom location.