Learn Morphome
What Morphome can do, what to upload, and how to get the best results — without reading documentation.
How it works
Upload an image
A photo, CAD drawing, sketch or 3D render. JPG, PNG, HEIC or PDF. Morphome detects what's in the image automatically.
Pick a direction
Use Visual Mode to set style and scene with one-click cards, run a Preset for an instant workflow, or switch to Free Mode and describe the result in your own words.
Generate & iterate
Download the result, or tweak the inputs and run again. Every generation is saved in your dashboard so you can compare or come back later.
What you can upload
Morphome works with five kinds of input. The system detects which one you uploaded and adjusts what it does next.
Photo
A real photograph of a building, room, object or scene. The system preserves geometry and restyles or recomposes around it.
CAD drawing
A technical line drawing, vector export or clean outline. Title blocks and dimensions are ignored — only the geometry matters.
Sketch
A freehand drawing, even rough. Proportions guide the result; the system fills in the photographic detail.
3D render
A SketchUp, Blender, Rhino, Revit or any clay/CGI render. Morphome materialises it into a photographic version of the same scene.
Aa
Just a description
No image needed. Switch to Free Mode and describe the scene fully — the model will produce it from your text alone.
What you can create
Morphome handles many subjects, but it's strongest in architecture and product visualisation.
Also handles landscapes, plants, people and generic scenes — useful as supporting elements, though the four areas above are where Morphome shines.
Two ways to control the output
Pick the workflow that matches how you think about the result.
Visual Mode
Default
Click-driven. You see your photo and pick from cards: Style, Background, Light, time of day, occupancy. Each option has a preview image, so you choose by looking — not by guessing English keywords.
Use it when
- You know what you want visually but not how to describe it
- You want consistent, predictable results
- You're iterating on the same image with different styles
Free Mode
For specifics
Write the result in your own words. Filters are off — you describe everything (style, light, mood, materials) in one paragraph. The uploaded image, if any, is used as reference only.
Use it when
- You have a specific look in mind the cards don't cover
- You want to add detail not exposed in Visual Mode
- You're generating from a text description with no input image
Presets — one-click workflows
A preset is a tuned mini-workflow: drop an image, hit run, get a polished result. No filter setup needed.
Make it Premium
Magazine-grade hero shot of any building. Same geometry, dramatically better lighting and finish.
Day cycle
Four matched frames: midday, sunset, night, dawn — same scene, four moments.
Four seasons
Same scene through spring, summer, autumn and winter — four photoreal frames.
Cinematic day to night
An 8-second timelapse from full daylight to night, with lights coming on around the scene.
Inputs that get great results
The model is forgiving, but a few habits make a real difference.
Sharp focus, good lighting
Photos with even light and crisp focus are read most accurately. Blurry phone shots produce blurry results.
Long side ≥ 1280 px
Bigger inputs give the model more detail to work with. Tiny thumbnails are harder to upscale convincingly.
Clean CAD with thin lines
Black-on-white outlines, no hatching or shadows. Title blocks and dimensions are ignored — leave them in if convenient.
One subject in frame
A single building, room or object reads more cleanly than a busy scene with many competing subjects.
Heavy watermarks
Diagonal watermarks across the image confuse the model and often leak into the result.
Extreme crops
Ultra-wide panoramas or tiny vertical slices truncate context. Standard 16:9, 4:3 or 1:1 framing is safest.
Pro tips
For photo inputs, leave the Background card untouched unless you specifically want a studio look. The model preserves real surroundings more faithfully than it invents new ones.
For CAD and sketch inputs of objects (a vehicle, a chair, a product), pick a studio background — white, grey, gradient. Outdoor scenes rarely make sense behind a schematic of a single object.
The "Add Prompt" panel is optional unless the task explicitly says Required. Most generations work fine with the cards alone.
If a result is close but not quite right, change one thing at a time. Changing style, light and background all at once makes it hard to tell what helped.
Video mode costs more than image mode and takes longer. Use it when motion matters — day-to-night transitions, season cycles, cinematic reveals — not as a default.
Every generation appears in your dashboard. Use it to compare variants side by side before settling on one.
Ready to try?
Upload an image and explore. Free credits are included on signup.