A fractal is a shape or pattern that repeats itself at every size. When you zoom
in, you keep finding smaller versions of the same pattern, almost like the shape
goes on forever. It is a never ending pattern that keeps looking similar no matter
how close you get.
Key points
- A fractal repeats its pattern again and again.
- Zooming in reveals smaller versions of the same shape.
- The pattern can go on forever in theory.
- Examples in nature include trees, ferns and coastlines.
For more detailed information on fractals view the
Fractals Wikipedia page.
Fractal Wallpapers
You can use this application to generate custom fractal wallpapers: explore with
the controls below until you find a fractal view you like, switch to fullscreen
mode, then use the screenshot button to download a full-size PNG that you can set
as your wallpaper.
Pan & Zoom Controls
- Pan: Click and drag on the fractal (mouse left button)
-
Select Area: Hold Shift and drag to draw a
selection box, then release to zoom to that area
-
Double-Click Zoom: Double-click anywhere on the fractal to zoom
in around that point
-
Precise Coordinate Zoom: Hold Ctrl to show
live coordinates under the cursor, then single-click to zoom into that exact
point
-
Scroll Wheel Zoom: Use the mouse wheel to zoom in and out
smoothly around the cursor
Fractal Selection
Choose from over 100 different fractal types using the dropdown menu. Fractals are
organized into families including Mandelbrot variants, Julia sets, Sierpinski
patterns, Koch curves, dragon curves, space-filling curves, attractors, and many
more.
Color Schemes
Experiment with 35+ color schemes to change the visual appearance of fractals.
Options include Classic, Fire, Ocean, Rainbow variants, Monochrome, Forest,
Sunset, and many artistic palettes. Each scheme creates a unique visual
interpretation of the same fractal structure. Access the Colour Scheme selector
and presets in the right panel (Themes & Presets).
You can also create your own custom color themes using the Color Scheme Editor
section in the right panel. Design unique gradients with multiple color stops,
save your creations, and even export/import palettes to share with others.
Iterations
Adjust the iteration count (10-400) to control the level of detail. Higher
iterations reveal more intricate patterns but take longer to render. The
application automatically adjusts quality to maintain smooth performance.
Export Options
-
Screenshot: Click the screenshot button to save the current
view as a PNG image
-
Share URL: Generate a shareable URL that encodes the fractal
state, allowing others to view the exact same fractal configuration
Machine Learning & Discovery
FractalAI includes an intelligent discovery system powered by machine learning
that learns from your preferences to find interesting fractal views automatically.
-
Favorites: Click the star icon to save interesting fractals to
your favorites collection. Your favorites are stored locally in your browser
-
ML Training: After adding 5 or more favorites, the system
automatically trains a neural network to learn your preferences. The model
analyzes features like fractal type, zoom level, position, and color scheme
-
Surprise Me: Use the "Surprise Me" button to discover new
interesting fractals. The system generates candidate views, validates them for
visual interest, and scores them using a hybrid algorithm combining fast
heuristics with ML-based refinement
-
Personalized Discovery: The ML model continuously learns from
your favorites, improving its ability to suggest fractals that match your
aesthetic preferences. The model retrains automatically in the background as you
add more favorites
-
Hybrid Algorithm: The discovery system uses a two-stage
approach: fast heuristic screening to filter out uninteresting views, followed
by ML-based scoring to rank the most promising candidates. This ensures both
speed and quality in discovery
-
View Favorites: Access your saved favorites from the favorites
panel. Click any favorite to instantly load that fractal configuration
The machine learning system uses Synaptic.js neural networks and stores both the
trained model and your favorites in browser local storage, so your preferences
persist across sessions.