Generate images with SUPA AI
SUPA is an AI-based tool that generates images from text prompts
-
Generate photorealistic images
We use the latest AI models to create high-quality, lifelike images.
-
Get inspired by community prompts
Discover new ideas in our library of community-generated images.
-
Try preset styles or create your own
You can choose from different visual styles to generate your masterpiece.
AI-generated image samples
Here are a few sample images generated using the SUPA AI
A rubber duck swimming in a bathtub in a medieval castle
A street orchestra playing Bach in the biggest stadium in Milano
A red bike parked under a tree in a rural area, with a beautiful field in the background
Corgis riding a bike
A car on a cyberpunk street
A business meeting in a conference hall
How to generate an image using SUPA
A step-by-step guide to generating pictures using the SUPA AI
Sign up for SUPA and confirm your email
After creating your account, go to your profile and create a new project with your preferred dimensions
In the left-side menu, click “AI”, then “Generate an image from a text prompt”
Type in your prompt and click “Generate”. You can also select an image style
FAQ
Yes, but you'll have a limited number of image generations. Subscribe to premium to generate more images
Yes, so long as the images are not in violation of anyone’s rights, you can use them freely. One exception is company logos that might be featured in a generated image. We recommend examining all AI-generated images before using them in your projects.
AI creates images by interpreting your text and turning it into visuals. You can describe an image you want to see, and AI will create it for you.
This technology is based on deep learning algorithms. AI scans through millions of texts and images to match a visual to its textual description. After that, it generates unique images based on your prompts.
Yes! You can detail your prompt by specifying the style, color palette, mood, setting, and even time period. The more specific your prompt is, the more the final image will match your vision.
Currently the AI only supports English.
How to write an efficient prompt
Decide on the purpose of the image
Before writing a prompt, ask yourself:
The clearer your goals are, the better your prompt will be.
- What is the purpose of this image? (e.g. advertising, illustration, concept-art, etc.)
- What visual elements are required to convey the idea?
- What feelings should this image evoke?
The clearer your goals are, the better your prompt will be.
Be specific
It’s crucial for the AI that your prompt includes specifics. For example, instead of using “dog”, try typing “big golden labrador dog sitting in a green field on a sunny day”. Add details: pose, expression, setting, texture, style.
Don’t: “rural house”
Do: “old wooden house with a thatched roof, surrounded by a blooming garden, on a sunny summer day, background is a blue sky with light clouds”
Don’t: “rural house”
Do: “old wooden house with a thatched roof, surrounded by a blooming garden, on a sunny summer day, background is a blue sky with light clouds”
Specify a style
Specifying a visual style can significantly change AI-generated results. Clarify the style of the image you want to generate. You can use any style you want, but here are some examples:
Example: “Digital illustration, steampunk, featuring zeppelins and steam engines, using warm brown and golden shades”
- Realistic (e.g. photorealistic, cinematic, etc.)
- Illustration (e.g. watercolor, pencil drawing, etc.)
- Sci-fi and fantasy (e.g. science fiction, steampunk, cyberpunk, etc.)
- Traditional art (e.g. Renaissance painting, Japanese print, etc.)
Example: “Digital illustration, steampunk, featuring zeppelins and steam engines, using warm brown and golden shades”
Add mood descriptors
To get the image to convey the right mood, try:
Example: “A mysterious twilight forest, with rays of moonlight shining through a thick fog, dark and mysterious atmosphere”
- Using atmospheric descriptors, such as: cozy, mysterious, romantic, sad, inspiring, etc.
- Specifying the lighting: daylight, sunset, neon lights, dim candlelight, etc.
Example: “A mysterious twilight forest, with rays of moonlight shining through a thick fog, dark and mysterious atmosphere”
Use keywords
The AI responds well to keywords that describe:
Example: “A minimalist interior with wooden furniture, close-up, soft diffused light, pastel”
- Angle: close-up, bird’s eye view, side view, etc.
- Color palette: warm or cool colors, pastel, black-and-white, etc.
- Dynamics: static scene, movement, wind, snowfall, etc.
- Texture: smooth, grainy, shiny, matte, etc.
Example: “A minimalist interior with wooden furniture, close-up, soft diffused light, pastel”
Exclude specific elements
If you want to avoid certain elements, specify that in your prompt. For example: “no people”, “no background”, “no text”.
Используйте структурированные шаблоны
To get consistent quality results from AI, you can structure your prompts this way:
Example: [Main object] + [Context/background] + [Mood/atmosphere] + [Style/color palette]
- Main object: the central focus of your image
- Context: where the object is located
- Style and mood: what the image should look like
- Technical details: angle, lighting, textures
Example: [Main object] + [Context/background] + [Mood/atmosphere] + [Style/color palette]
Experiment and improve
You probably won’t get the perfect prompt on your first try — and that’s okay. Try to:
- Change the order of words
- Add or eliminate details
- Check how those changes are reflected in the generated images