NEWKeep editing in Agent Workspace.

Open Agent

Tool Directory

Image Generator

Swipe or tap to switch

Image Conditioned Generation

Image Conditioned Generation
Reference-Guided Variations

Generate reference-guided product variations, scene updates, and campaign-ready edits from one source image with optional extra references for stronger visual consistency.

REQUIRED

Examples: keep the product shape, place it on a clean studio background

0/500

Reference Images

Add up to 3 extra references to guide style, product details, or brand consistency.

Optional. Use reference images when you want stronger visual consistency across lighting, material, packaging, or product styling.
Guided workflow
Secure SSL

Recent Sessions

Loading Image Conditioned Generation sessions...

Reference-guided outputs

Maintain structure while generating new styles or scenes from one main image plus optional extra references.

Prompt control

Direct background swaps, colorways, and material updates.

Multi-reference guidance

Use extra references to reinforce art direction, details, and brand consistency.

Layer-ready handoff

Send the best outputs to Image-to-Layers for editing.

The Workflow

Generate Variations in 4 Steps

Upload a reference image, guide the prompt, review variations, and export for edits.

Step 1

Upload a reference image

Upload one clear main image so the model can preserve shape, framing, and key product details.

Step 2

Add optional extra references

Attach up to three more reference images when you want stronger consistency for styling, material cues, packaging details, or brand look-and-feel.

Step 3

Describe the change

Write a concise prompt that explains what should change, such as the setting, colorway, material, or lighting direction.

Step 4

Review and export

Compare the outputs for edge quality, brand accuracy, and scene consistency, then open the best result in Image to Layers for further cleanup or handoff.

Key Features

Reference guidance, controlled variations, and batch-ready outputs for ecommerce and marketing.

Variation generation

Create multiple candidates for ads and catalog testing.

Background swaps

Replace environments while keeping the product consistent.

Brand consistency

Preserve logos, proportions, and product geometry.

Batch workflows

Move faster on recurring catalog and campaign updates.

Layer handoff

Export to Image-to-Layers for PSD-ready edits.

50 credits
per generation
Image+Prompt
conditioned input
Batch-ready
catalog workflows
Reference-led
variation workflow

Common Use Cases

Generate product variations, marketing creatives, and scene swaps from a single reference image.

Product variations

Generate colorways and material changes from a single shot.

Marketing creatives

Produce ad variants while preserving product structure.

Lifestyle scenes

Swap backgrounds for seasonal or campaign visuals.

Catalog refresh

Update listing imagery without a reshoot.

Design exploration

Test visual directions before production.

Model + Credits

Use one reference image and one prompt to generate controlled visual variations.

Workflow
Best used for
Reference-guided product scenes, colorway updates, and campaign-ready variants.
Credit Cost
Each generation consumes
50 credits

Case Studies

Product catalog refresh

Challenge: New colorways needed without reshoots or delays.

Solution: Used one hero photo as the visual anchor, then generated controlled variations for each new finish and scene.

Result: 30% faster SKU expansion with consistent imagery.

Marketing creative testing

Challenge: Ad variants were slow to produce for A/B tests.

Solution: Built multiple on-brand directions from the same product shot so the team could compare concepts before launch.

Result: Faster iteration and stronger campaign learnings.

Lifestyle background swaps

Challenge: Seasonal scenes needed while preserving product structure.

Solution: Used the original packshot as the reference while changing only the setting, lighting mood, and styling cues.

Result: On-brand scenes without re-shoots.

Design exploration

Challenge: Teams needed multiple style directions before production.

Solution: Generated several visual directions from the same source image to align faster on the right creative direction.

Result: Shorter decision cycles and clearer direction.

Ecommerce scene generation

Challenge: Marketplace listings lacked consistent scene styling.

Solution: Created a repeatable reference-guided workflow for listing imagery so each product family kept a consistent look.

Result: Improved listing quality and cohesion.

Known Limitations

Style drift

Long prompts can shift the subject away from the reference.

Logo fidelity

Small text or logos may need manual review.

Complex lighting

Strong reflections can be inconsistent between variants.

Busy backgrounds

Dense scenes reduce structural consistency.

Low-resolution inputs

Pixelated sources reduce edge quality.

Frequently Asked Questions

What is image conditioned generation?

It creates new visuals from a source image and a prompt, helping you preserve composition while changing the surrounding scene or styling.

How is it different from text-only generation?

Text-only generation starts from scratch. Image conditioned generation keeps more of the source image structure so changes stay closer to the original asset.

Can I keep logos and proportions?

Yes, although you should still review the final output closely when logos, labels, or small text must remain exact.

What inputs are supported?

Clean JPG, PNG, and WebP files work best, especially when the subject is well lit and clearly separated from the background.

Can I change only the background?

Yes. This workflow is a strong fit for background swaps, scene styling changes, and contextual updates around the same subject.

Is it good for ecommerce?

Yes. It is useful for product variations, catalog refreshes, seasonal campaigns, and creative testing from one source photo.

How many variations should I generate?

Start with a small batch of 3 to 6 outputs, compare the strongest directions, then iterate on the best one.

Does it support batch workflows?

Yes. Teams often use it in repeatable catalog and campaign workflows where each product needs the same style treatment.

What file formats do I get?

Outputs are typically returned as PNG or JPG. You can move the result into Image to Layers if you need a more editable production workflow.

Is there an API?

This page is designed for the browser workflow, while teams with larger pipelines can adapt the same process to their own automation stack.

When should I use Image-to-Layers?

Use Image to Layers when you want to separate foreground elements, refine shadows, or hand the result off for layered editing.

How do I reduce artifacts?

Use clean references, keep prompts specific, and iterate in small steps so the model stays close to the source composition.

Ready to generate variations?

Create image-to-image variants and hand off the best outputs for layered editing.

Open Image to Layers
    Image Conditioned Generation | Reference-Guided Product Variations | Image to Layers