Use a clear pet photo and a separate scene or costume reference to create a new portrait while telling the model which markings, features, and proportions must stay recognizable.
Daily credits are always used first. Files expire after 24 hours.
Explore AI Creations
Inspiration examples, not your generated result.
Inspiration examplePortrait + Fantasy
See how two references can guide one fantasy character.
How to use it
A pet AI generator in this workspace does not paste a costume over a photo. It treats the pet as the identity reference and the second image as creative context, then generates a new composition. The result can vary between runs, so important fur patterns and anatomy still need a careful review.
Put a sharp pet photo in Image 1, then upload one scene, costume, lighting, or palette reference as Image 2. Keep eyes, ears, muzzle, coat markings, and the pet's outline visible.
Tell the model that Image 1 protects the pet's identity and Image 2 supplies one creative direction. Name the eye color, ear shape, coat pattern, collar, or proportions that must survive.
Generate the new portrait, then check the face, fur pattern, paws, tail, costume edges, scale, and ground contact. Download only after the pet still reads as the intended subject.
Why use two references
These four practical principles explain why the pet AI generator gives each image one clear job before it generates a new composition.
Image 1 supplies visual evidence such as muzzle shape, ear position, coat pattern, eye color, and body proportions. The prompt should name the two or three traits that matter most instead of asking the model to preserve everything equally.
Image 2 should contribute one coherent job: a location, costume, lighting setup, palette, or illustration direction. It is guidance for a new scene, not a layer pasted over the original pet photo.
A relationship-first prompt explains which pet details must survive and which environmental details may change. Compatible camera angles and simple poses reduce the chance of distorted paws, duplicated tails, or clothing that hides the face.
Inspect the face, coat markings, ears, limb count, tail, scale, ground contact, and any costume edge. Regenerate or correct one concrete problem at a time rather than assuming the same inputs will always produce the same result.
Six ideas to try
Each story keeps the user goal, both input roles, and generated result visible. Open the details for the prompt goal, review checklist, failure risk, and recorded provenance.
A recognizable pet becomes the single subject of a playful space portrait instead of being pasted over a costume reference.
Project-owned synthetic demonstration generated with OpenAI; not a customer result, live benchmark, or repeatability promise. Provenance: pet-astronaut-triptych.
A focused environment reference changes the setting and light while the cat's face and chest pattern remain reviewable.
Project-owned generated demonstration created with OpenAI; not a user commission, customer result, or benchmark. Provenance: pet-greenhouse-triptych.
A short-legged corgi is moved into a winter cabin scene while its ears, blaze, and compact proportions stay explicit.
Project-owned synthetic demonstration generated for this site; output details can vary and require visual review. Provenance: pet-corgi-alpine-triptych.
A minimal coastal reference tests whether a long, lean silhouette can survive a stronger editorial change in light and setting.
Project-owned synthetic demonstration generated for explanation; not a customer photograph or breed-accuracy benchmark. Provenance: pet-greyhound-coast-triptych.
A small rabbit and low meadow reference show why scale, ear shape, and foreground placement need precise instructions.
Project-owned generated demonstration; it explains reference roles and does not promise the same animal identity on every run. Provenance: pet-rabbit-meadow-triptych.
A colorful bird example makes feather boundaries, perch contact, and separation from a similarly colored environment easy to inspect.
Project-owned synthetic demonstration; not a wildlife photograph, customer result, or claim of zoological precision. Provenance: pet-parrot-conservatory-triptych.
Prompt starters
Open one starting point, then replace its preservation and change instructions with details from your own two references.
Keep the pet from Image 1 recognizable, including its face, coat pattern, and proportions. Use the costume and setting from Image 2 as inspiration, fitted naturally around the animal without hiding its eyes or muzzle.
Place the pet from Image 1 into the environment shown in Image 2. Match the direction of light and ground contact while preserving realistic anatomy and the pet's distinctive markings.
Create an original illustrated portrait using the pet in Image 1 as the identity reference and the palette and texture of Image 2 as visual direction. Do not imitate a named artist or franchise.
Review limits
Generation is variable. Use this compact review list when a result drifts, distorts a detail, or overuses the second reference.
Creator perspectives
Illustrative creator story — not a verified customer testimonial. These first-person scenarios show how different creators might describe a two-reference project.
I wanted to explore a calm keepsake of a much-missed pet, so I used the first image for the greyhound's long silhouette and the second only for soft coastal light.
Pet remembrance concept
Illustrative creator story
I imagined a playful gift portrait, using the pet photo to protect the retriever's face and the astronaut reference to guide the costume and cabin setting.
Creative gift concept
Illustrative creator story
I planned a colorful pet-content image and kept the parrot as the single subject while the conservatory reference supplied only foliage, light, and atmosphere.
Pet content concept
Illustrative creator story
Related workflows
Use a route whose inputs match the image you actually want to make, or read the practical two-reference guide before generating.
Frequently asked questions
