When is GPT Image 3 Coming Out? Latest News and Rumors

There’s been growing speculation about OpenAI GPT Image 3 after users spotted three unannounced “Aurora” image models in OpenAI’s API and Arena tests. Some believe these could be early versions of OpenAI’s next image model, but OpenAI has yet to confirm or announce GPT Image 3.

So, what’s really going on with the Aurora models? Is the GPT Image 3 release date closer than we think? We’ve gathered all the clues behind these leaks, what improvements the next model might bring, and everything you need to know to stay up to date.

Part 1. About GPT Image

GPT Image is OpenAI’s family of AI models for generating and editing images. It powers ChatGPT Images, where you can create pictures from text prompts, upload existing images for edits, and refine results through follow-up instructions. Developers can also access GPT Image models through OpenAI’s API.

The GPT Image series has come a long way since OpenAI first introduced image generation through DALL·E. Recently, speculation about GPT Image 3 release has caught the attention of AI enthusiasts eager to see what OpenAI might release next. Even so, OpenAI hasn’t confirmed that it is developing under this name, and there is no official feature list or release date yet.

GPT Image 3 Release Date: What Do We Know?

There is no confirmed date for when GPT Image 3 is coming out. The current speculation mainly comes from reports of three unidentified image models spotted in Arena tests.

On October 18, 2026, X user @Angaisb_ reported seeing three new image models grouped under the codenames “Aurora Trio” (aurora-red-alpha, aurora-green-alpha, and aurora-blue-alpha) in Arena, an AI model testing platform. The discovery sparked speculation that they could be early versions of GPT Image 3.

However, the Aurora model references reportedly disappeared from OpenAI’s API list shortly after they were spotted. The sightings also don’t confirm that these models belong to OpenAI’s next image generation release.

Part 2. The Launch of GPT Image 2.6

While many users suspect the Aurora Trio could be early versions of GPT Image 3, some also suggest a possibility that these models are related to GPT Image 2.6 or another incremental update.

ChatGPT Image 2.6 is the latest image generation model by OpenAI. It was released on November 4, 2026, amid growing speculation about GPT Image 3. The update brings faster image generation, more consistent editing, and a new GIF creation feature.

But after some digging, there’s no confirmed connection between the Aurora Trio and GPT Image 2.6. We can only wait for OpenAI to share more details before drawing any conclusions.

GPT Image 3 vs GPT Image 2.6 vs GPT Image 2.5

While GPT Image 3 is still not confirmed, we can expect it to improve on the two latest versions, GPT Image 2.6 and GPT Image 2.5.

Note: The comparison is based on OpenAI’s release announcements. GPT Image 3 entries are predictions based on the direction of previous updates, not confirmed specifications.

Part 3. What to Expect from GPT Image 3 Features

Since OpenAI has not confirmed GPT Image 3.0 or shared any details about its features, there's no official feature list to refer to yet. However, based on the improvements in previous releases and current developments in AI image generation, we can make a few guesses about what the next version might offer:

• More accurate text in images. Text rendering has improved across recent image generators, but it can still be difficult to get every word, letter, and small detail right. GPT Image 3 could take this further with more accurate text placement and fewer spelling mistakes.
• More precise image editing. GPT Image 2.5 and 2.6 already introduced improvements to editing accuracy and consistency. GPT Image 3 could build on these features by following more specific instructions while leaving unrelated parts of an image untouched.
• Better understanding of complex prompts. There’s a possible improvement in stronger instruction following. A more advanced model could handle prompts with several subjects, specific object positions, lighting requirements, and detailed background elements more reliably.

Is GPT Image 3 Free?

GPT Image 3 may be available to free ChatGPT users if OpenAI follows the same approach as it did with previous image models. For now, the latest ChatGPT Images 2.6 is available across all subscription tiers, including the Free plan, though usage limits may vary.

The same applies to API access. There’s no telling whether GPT Image 3 API will be available or how much it might cost.

Part 4. What the AI Community Thinks About GPT Image 3

Reddit and X have become places where AI enthusiasts, hobbyists, and regular ChatGPT users share their thoughts about the rumored GPT Image 3. Some are excited about the possibility of a new image model, while others remain skeptical about how much it could improve.

Many hope GPT Image 3 will address some of the issues they have noticed in earlier versions. Common expectations mentioned in GPT Image 3 Reddit discussions include:

• More realistic image quality
• Fewer visual artifacts
• Better text rendering
• More accurate prompt understanding
• Consistent characters across multiple images

Better editing controls are also high on the wish list, especially the ability to change one part of an image without affecting the rest.

Bonus: Generate with GPT Image and Turn Them into Videos with Filmora

AI-generated images can be more than just still pictures. While waiting for GPT Image 3, you can already create images with GPT Image 2 and turn them into videos using Wondershare Filmora.

Filmora offers a range of AI tools, including an AI Image Generator powered by GPT Image 2 and other advanced models. You can create images from text prompts, then turn them into videos and add your own creative touches to make something worth sharing.

Go Beyond Image Generation with Filmora

For video creation, Filmora offers several AI features that can help you get started:

• Image-to-Video: Turn your AI-generated images into video clips with AI-generated motion.
• AI Effects Library: Apply ready-made effects and animations to your clips to create a specific visual style.
• Text-to-Video: Create video clips from text prompts without needing an existing image.

All these tools are built into a video editing workspace with a timeline, where you can arrange clips, add music and effects, and make further adjustments. Its interface is also intuitive and simple enough to get started without much editing experience.

Try It Free Try It Free

Download the APP

Scan to get the Filmora App

Install free Filmora App Install free Filmora App

Secure Download

Conclusion

Despite the growing rumors about GPT Image 3, OpenAI has yet to make an official announcement. Its release date, expected features, and API availability therefore remain unknown. The recent launch of GPT Image 2.6 gives us a good idea of how the technology continues to improve, but it’s still too early to tell what the next major version will bring.

For now, there's no need to wait for GPT Image 3 to explore AI image creation. GPT Image 2.6 is already available in ChatGPT, while tools like Filmora let you take AI-generated images further with video generation and editing features.