Generate images with structure control
Upload a photo or sketch, pick depth or edge mode and describe what you want.
By using this tool you agree to our Terms of Service. Only upload images you have the rights to.
A failed job costs nothing — credits are returned automatically. We keep photo and video results for 30 days, and guest files are deleted after 14 days.
Credit packs from $1.49, one-time. Credits never expire and no subscription is required.
AI image generation
FLUX Depth Pro or FLUX Canny Pro generates a new image with the layout and shapes of your reference. Requires a Starter plan or higher.
- 2 modes: depth and edges
- Starter plan or higher
- 3 credits per image
- Operational
- 3 cr · FLUX Depth/Canny Pro
How it works
- Step 1
Upload a reference
A photo or sketch as JPG, PNG or WEBP, up to 20 MB.
- Step 2
Pick a mode
Depth or edges.
- Step 3
Describe the result
A description is required: style, materials, light, mood.
- Step 4
Generate and download
One generation gives one PNG image.
What is Sketch → image (same layout)?
Sketch → image (same layout) is an AI tool in Pixelift that runs in the browser with no installation. Upload a photo or sketch, pick depth or edge mode and describe what you want. FLUX Depth Pro or FLUX Canny Pro generates a new image with the layout and shapes of your reference. Requires a Starter plan or higher. Engine: FLUX Depth/Canny Pro. Cost: 3 credits per result; credits are bought once in packs from $1.49, no subscription, and they never expire. Files go to our server in the EU; if the tool uses an AI model, it runs at a provider from our subprocessor list. We never train models on them.
Two control modes
Depth control
FLUX Depth Pro reads the depth of your reference and keeps the space: what is near and what is far.
Edge control (Canny)
FLUX Canny Pro follows the edges and outlines of your reference, so shapes stay in place.
Why structure control
Layout from your reference
Composition and shapes come from your photo, not from chance.
New content from your description
Describe style, materials, light and mood in words; the model expands your description before generating.
No sliders
Generation settings are fixed. You steer the result with the mode and your description.
Made for creators
Concept art
A sketch or photo as the skeleton of a finished illustration.
Product design
Product variants in new materials and colors with the same shape.
Architecture
A building in another style or season, with the same massing and perspective.
Interiors
The same room with a new design, keeping its layout and perspective.
Technology: FLUX Depth Pro and FLUX Canny Pro
We use two Black Forest Labs models on Replicate: FLUX Depth Pro guides generation with a depth map and FLUX Canny Pro with an edge map, both extracted from your reference.
- Depth map extracted from the reference image
- Canny edge detection
- Two control modes: depth or edges
- Fixed generation settings, no sliders
What to expect — ✓ Works best for
- Images with clear structure and depth
- Architecture and interior scenes
- Objects and sketches with readable shapes
Limitations
- Flat images give weak depth maps
- Complex scenes need a refined description
- There is no pose mode and no strength slider
Tips for best results
- Use depth mode for architecture and interiors
- Edge mode works best for graphics, logos and sketches with clear outlines
- Write a detailed description of style, materials and light
- Result too strict or too loose? Change the mode or the description, as strength cannot be set
- Use sharp references with good contrast
- Compare both modes on the same photo
Frequently asked questions
What is the difference between the modes?
Depth keeps spatial relationships, meaning what is near and what is far. Edges keep outlines and shapes. There is no pose control.
Which mode should I choose?
Depth for interiors, architecture and landscapes; edges for graphics, logos, sketches and products.
Is the content of the reference image preserved?
No, only the structure: depth or edges. Content, style and colors come from your description.
Can I set how strongly the reference is followed?
No. Generation settings are fixed. You change the result with the mode and your description.
How much does it cost and who can use it?
Each generation costs 3 credits and requires an active Starter plan or higher. Credits are returned if generation fails.
What files can I upload?
JPG, PNG or WEBP up to 20 MB. The result is a PNG file.
What to Do Next
Continue your workflow with these tools
More tools
Create from a prompt
- Popular
AI Image Generator
Type what you want to see — get a finished image in seconds.
0.5-7 kr
Custom Emoji
Your own emoji from a short description.
0.5 kr
- Popular
Upscale & Sharpen
A small or soft photo, large and crisp, up to 8×. Usually 17 seconds.
1-4 kr
Colorize Old Photo
Black-and-white photos get natural color. Usually about 40 seconds.
1.5 kr
Remove Noise & Grain
Clean up noisy, grainy or over-compressed photos.
0.5-6 kr
Start generating with structure control
The tool works on the Starter plan and higher. See the plans on the pricing page.
Call it from code
One endpoint, key in the header. The same tool that runs on this page — no browser, no clicking.
curl -X POST https://pixelift.pl/api/structure-control \
-H "Authorization: Bearer pk_live_..." \
-F "file=@zdjecie.jpg" \
-F "prompt=nowoczesne wnetrze" \
-F "control_mode=depth" \
-F "strength=0.7"Response: { "processedImage": "data:image/png;base64,iVBORw0…", "creditsUsed": 1 }
Request fields
file | file | required | Obraz referencyjny |
prompt | string | required | |
control_mode | depth · canny · pose | optional | |
strength | number | optional |