Skip to main content

Image Generation & Editing

Create stunning images and edit existing ones using AI-powered tools. Bike4Mind integrates multiple image generation providers to give you flexibility and creative control.

Generating Images​

How to Generate​

Using the Model Picker (Recommended):

  1. Click the model picker in the chat input area
  2. Select Image Models from the dropdown filter
  3. Choose an image model (e.g., FLUX PRO, GPT-Image-1)
  4. Type a description of the image you want
  5. Press Enter to generate

Using the Command:

You can also use the /gen_image command followed by your prompt:

/gen_image A serene mountain landscape at sunset with a lake reflection

Once the model generates your image, you can view, download, or edit the result.

Available Models​

OpenAI (DALL-E / GPT-Image)​

  • GPT-Image-1: Latest OpenAI image generation
  • Quality options: Standard or HD
  • Sizes: 1024x1024, 1024x1536, 1536x1024

Flux (BlackForest Labs)​

  • FLUX PRO: High-quality standard generation
  • FLUX PRO Ultra: Premium quality with custom aspect ratios
  • FLUX Kontext: Transform existing images based on prompts

Generation Settings​

SettingDescription
QualityStandard or HD (affects detail and credits)
Size/Aspect RatioChoose dimensions for your image
Number of ImagesGenerate 1-4 images per request
Safety LevelContent moderation sensitivity (Flux models)

Editing Images​

Inpainting (Edit Parts of an Image)​

Edit specific areas of an existing image while keeping the rest intact:

  1. Click Edit on any generated image
  2. Use the brush tool to paint over areas you want to change
  3. Describe what you want in those areas
  4. Click Generate to apply the edit

Image-to-Image Transformation​

Transform an entire image based on a new prompt:

  1. Upload or select an existing image
  2. Describe how you want it transformed
  3. The AI will create a new version based on your description

Masking Tools​

  • Brush size: Adjust from 1-100 pixels
  • Undo/Redo: Fix mistakes easily
  • Clear mask: Start over if needed
  • Preview: See your mask before generating

Working with Generated Images​

Image Actions​

Every generated image includes these options:

ActionDescription
DownloadSave to your device
CopyCopy to clipboard
EditOpen in the image editor
ZoomView at full resolution
GalleryBrowse multiple images

When multiple images are generated:

  • Use arrow keys or buttons to navigate
  • Click thumbnails for quick selection
  • Download individual or all images

Tips for Better Results​

Writing Effective Prompts​

  1. Be specific: "A golden retriever playing in autumn leaves at sunset" works better than "a dog"
  2. Include style: Mention artistic styles like "watercolor," "photorealistic," or "digital art"
  3. Describe composition: Specify "close-up," "wide shot," or "from above"
  4. Add atmosphere: Include lighting, mood, and environment details

Common Prompt Elements​

  • Subject: What's the main focus?
  • Style: Photorealistic, cartoon, oil painting, etc.
  • Lighting: Dramatic, soft, golden hour, studio
  • Mood: Peaceful, energetic, mysterious
  • Details: Colors, textures, specific features

Credit Usage​

Image generation uses credits based on:

  • Model: Different models have different costs
  • Quality: HD images cost more than standard
  • Size: Larger images use more credits
  • Quantity: Each image in a batch counts separately

Check your credit balance in Profile > Credit Analytics.

Troubleshooting​

Image Not Generating​

  • Make sure you have an image model selected (use the model picker and filter by Image Models)
  • Check you have sufficient credits
  • Verify you have API keys configured for the image provider (OpenAI or BlackForest Labs)
  • Try a simpler prompt first

Poor Quality Results​

  • Add more detail to your prompt
  • Try a different model
  • Use HD quality setting
  • Specify the style you want

Edit Not Working as Expected​

  • Make sure your mask covers the right areas
  • Be specific about what should appear in masked areas
  • Try a smaller edit area first