GPT Image 1.5 AI Photo Editor and Image Generator

OpenAI's GPT Image 1.5 makes precise, everyday edits easy: swap a background, try on a new outfit, or recolor a product while faces, logos, and lighting stay the same. Generate from text or edit with up to 16 reference images—at the lowest price of any OpenAI model on PixPal.

💡 Tip: say what to change and what to keep, like "change the jacket to red, keep her face and the background"

Edits that keep the rest

Change the background, an outfit, or a color, and the lighting, composition, and faces you didn't mention stay put.

Faces and logos that hold

People and brand marks stay recognizable across multiple rounds of edits, which matters for marketing and product work.

Up to 16 reference images

Combine a person, a product, a place, and a style reference in one request and tell the model what each one is for.

Lowest OpenAI price on PixPal

At default settings, GPT Image 1.5 uses fewer credits per image than GPT Image 2 or 2.5, so everyday edits and quick tests stay affordable.

Three steps with GPT Image 1.5

Step 1

Add a reference

Upload a product photo, portrait, or mood board—or start from text only.

Step 2

Describe your vision

Describe the layout, style, lighting, and props. GPT Image 1.5 understands long, detailed prompts.

Step 3

Generate & export

Pick your favorite result and download high-resolution files ready for stores and ads.

Introduce

What is GPT Image 1.5?

GPT Image 1.5 is an image generation and editing model from OpenAI. It launched on December 16, 2025, alongside the redesigned ChatGPT Images, and OpenAI described it as a shift from novelty image generation to practical, everyday visual work. It was OpenAI's flagship image model until GPT Image 2 arrived in 2026.

Compared with GPT Image 1, OpenAI says it follows instructions more closely, edits more precisely, and generates images up to four times faster in ChatGPT. The biggest improvement is in editing: when you change one thing, it does a better job of keeping everything else—lighting, composition, a person's face, and brand logos—the way it was. That makes it well suited to clothing and hairstyle try-ons, background swaps, and product catalogs built from a single photo. It also renders dense and small text better than its predecessor.

On PixPal, GPT Image 1.5 is the most affordable OpenAI model in the studio. You can write a prompt to create a new image, or upload up to 16 reference images to edit or combine them. Choose a square, portrait, or landscape size up to 1536 pixels on the long side, then pick Low, Medium, or High quality.

What you can do with GPT Image 1.5

Practical edits and everyday images—without learning complex photo-editing software.

Before and after photo of a woman with the background changed to a beach using GPT Image 1.5

Everyday photo edits

Replace a messy background with a beach, a studio, or a clean office; remove clutter; or change the time of day. Upload your photo, describe the change in plain language, and GPT Image 1.5 keeps the person, the pose, and the lighting consistent with the original.

The same model shown in a bathrobe and a silk top, an outfit try-on made with GPT Image 1.5

Outfit and hairstyle try-ons

See yourself or a model in a different jacket, dress, or haircut before you buy or shoot. OpenAI specifically improved clothing and hairstyle try-ons in GPT Image 1.5, so the face and body stay true while only the look changes.

One backpack shown in four colors and scenes, a product catalog made with GPT Image 1.5

Product catalogs from one photo

Start with a single product shot and create color variants, lifestyle scenes, and new angles for Shopify, Amazon, or Etsy listings. GPT Image 1.5 keeps logos, stitching, and shapes consistent, so every image in the catalog looks like the same product.

A stack of social media graphics and a phone showing a feed, made with GPT Image 1.5

Social posts and simple graphics

Make Instagram posts, blog headers, thumbnails, and simple promo graphics at a low cost per image. Because GPT Image 1.5 handles small and dense text better than earlier OpenAI models, short captions and labels are more likely to come out clean—just proofread before posting.

GPT Image 1.5 features and settings

Everything you can control in the PixPal studio.

Text to image

Describe a scene to create portraits, product shots, social graphics, and illustrations.

Image to image

Edit a photo, try on outfits, or restyle a scene while keeping faces and products recognizable.

Up to 16 references

Upload JPEG, PNG, or WebP images up to 10 MB each and say what each one should contribute.

Three sizes

Square, portrait, or landscape: 1:1, 2:3, 3:2, or exactly 1024×1024, 1024×1536, or 1536×1024.

Low, Medium, and High quality

Medium is the default. Use Low for quick tests and High for fine detail.

Better small text

Short captions, labels, and dense text render more cleanly than with GPT Image 1.

Reference photos of people, products, and places merged into one scene with GPT Image 1.5

Combine up to 16 references

Upload a portrait, a product, a location, and a style sample in the same request, and GPT Image 1.5 blends them into one image. Label each one in your prompt—"image 1 is the model, image 2 is the bag, image 3 is the setting"—so the model knows which details to keep from which photo.

Blue fabric texture at Low, Medium, and High quality in GPT Image 1.5

Three quality levels, one low price

GPT Image 1.5 is the least expensive OpenAI model on PixPal. Use Low to test ideas quickly, Medium (the default) for most everyday work, and High when you need fine textures like fabric, hair, or skin. Higher quality takes a little longer but brings out more detail.

Portrait, skincare product still life, and mountain lake photos made with GPT Image 1.5

Realistic people, products, and places

From natural-looking portraits to clean product shots and landscapes, GPT Image 1.5 produces believable photographs and handles illustration styles when you ask for them. Describe the lighting, lens, and mood the way a photographer would for the most natural results.

What creators are saying

I shot one photo of our backpack and used GPT Image 1.5 to make it in four colors and three settings. The zippers and straps looked the same in every image.

Maya Chen

Commercial photographer

I asked GPT Image 1.5 to swap the cluttered background in my headshot for a clean studio. My face and hair didn't change at all.

Leo Park

Social media lead

Before I buy a jacket online, I try it on myself with GPT Image 1.5. It's surprisingly close to how it looks in real life.

Elena Voss

Creative director

For quick blog headers and Instagram posts, GPT Image 1.5 is good enough and costs less per image than the newer models.

Aisha Rahman

DTC brand owner

GPT Image 1.5 FAQ

Everything you need to know about using OpenAI GPT Image 1.5 on PixPal.

Pair GPT Image 1.5 with these tools

Object RemoverOutfit SwapImage to PromptCreative TemplatesBackground RemoverImage UpscalerAI Expand Image

Start editing with GPT Image 1.5

Swap backgrounds, try on outfits, and build product photos with OpenAI's everyday image model—right in your browser.