OpenAI's most advanced image generation model with precise editing, 4x faster generation, and exceptional instruction following. Create and transform images with unmatched fidelity.
Released December 16, 2025 — OpenAI's most advanced image generation model to date
Change only what you ask for while preserving lighting, composition, and facial likeness across edits with exceptional consistency.
Generate images up to 4 times faster than previous versions, making iteration and creative exploration effortless.
Handle denser and smaller text with improved accuracy, perfect for infographics, posters, and detailed compositions.
Follow complex instructions more reliably, enabling intricate compositions where relationships between elements are preserved as intended.
Real outputs generated with this model — every caption is the exact prompt used.
A bold mid-century travel poster — 'ESCAPE NORTH', a rippling aurora over geometric fjords in teal, coral and cream.
A studio e-commerce shot — a frosted-glass perfume bottle with a brushed-gold cap on wet black stone, dewy droplets and a soft mirror reflection.
A precisely arranged top-down knolling flat-lay on oak — leather journal centered, brass compass upper-left, film camera bottom-center, all aligned to a neat grid.
A high-fashion editorial portrait — a structured emerald wool coat on a warm terracotta backdrop, soft beauty-dish light and immaculate fabric detail.
A dreamy album-cover illustration — a lone sailboat on a paper-cut dusk sea beneath a giant crescent moon in indigo, plum and gold.
A clean flat-design infographic — 'A Cup of Coffee' with labeled callouts for Water, Crema, Espresso and Aroma in a muted brown-and-teal palette.
Write your own prompt and generate high-resolution images with GPT Image 1.5. Start free — no credit card required.
Try GPT Image 1.5GPT Image 1.5, released on December 16, 2025, represents OpenAI's most advanced image generation model to date. Built on cutting-edge multimodal language model technology, it delivers high-fidelity images with exceptional prompt adherence and semantic understanding. The model excels at preserving composition, lighting, and fine-grained details while enabling both practical edits and expressive transformations.
Designed as a creative studio in your pocket, GPT Image 1.5 combines powerful text-to-image generation with sophisticated editing capabilities. Whether you're making subtle adjustments or complete transformations, the model maintains consistency across inputs, outputs, and subsequent edits—all while generating results up to 4 times faster than its predecessors.
Generate stunning images from text prompts with exceptional detail preservation, accurate composition, and natural lighting. The model understands complex scenes and relationships between elements — delivering strong prompt adherence, natural-looking results, and fine-grained detail preservation.
Edit uploaded images with surgical precision, changing only what you specify while maintaining consistency in lighting, composition, and facial features across multiple edits. Supports adding, subtracting, and combining; blending and transposing; and facial likeness preservation.
Apply expressive transformations that change style, add elements, and reimagine concepts while retaining the essence of the original image. Perfect for artistic exploration, with stylistic filters, conceptual transformations, and preset styles available.
Visualize clothing, hairstyles, and accessories on people with believable results. The model maintains facial features and body proportions while applying changes — enabling clothing visualization, hairstyle experimentation, and consistent appearance.
Render complex text layouts with improved accuracy, including small fonts, dense paragraphs, and structured content like infographics and code snippets. Handles markdown rendering, infographic creation, and code visualization.
Experience up to 4x faster generation compared to previous models, enabling rapid iteration and seamless creative workflows without long wait times. Delivers quick iteration cycles, real-time exploration, and reduced waiting time.
| Resolution | 1024x1024, 1024x1536, 1536x1024 |
|---|---|
| Quality Levels | Low, Medium, High |
| Format | PNG, JPEG |
| Aspect Ratios | 1:1, 2:3, 3:2 |
| Model Name | gpt-image-1.5 |
|---|---|
| Release Date | December 16, 2025 |
| Performance | Highest (5/5) |
| Speed | Medium (4x faster) |
| Text-to-Image | Yes |
|---|---|
| Image Editing | Yes |
| Multi-Turn Editing | Yes |
| Batch Generation | Yes |
Edit photos with precision while maintaining lighting, composition, and facial features. Perfect for portrait retouching, background changes, and object removal.
Create product visualizations, virtual try-ons, and fashion lookbooks. Generate lifestyle images and model different clothing combinations efficiently.
Design eye-catching ads, social media content, and promotional materials. Create multiple variations quickly for A/B testing and campaign optimization.
Generate movie posters, album covers, book illustrations, and concept art. Apply creative transformations and stylistic filters for artistic projects.
Create professional infographics with dense text and complex layouts. Render charts, diagrams, and educational materials with accurate text rendering.
Rapidly prototype UI designs, product mockups, and brand identities. Iterate on concepts quickly with fast generation and precise editing capabilities.
While the model handles facial consistency well, generating images with many distinct faces (10+) in complex arrangements may show reduced accuracy in maintaining individual features.
Some highly specific artistic styles or niche visual aesthetics may not be perfectly replicated. The model works best with common styles and mainstream artistic directions.
While English text rendering is excellent, non-Latin scripts and complex multilingual text layouts may have reduced accuracy, especially for small font sizes or dense text.
The model may occasionally generate physically implausible details, such as incorrect hand anatomy, impossible reflections, or inconsistent perspectives in complex scenes.
Experience the most advanced image generation and editing model from OpenAI. Create, transform, and refine images with unprecedented precision and speed.
Try GPT Image 1.5 now