Images produced by ChatGPT show remarkable progress: two pictures generated with the same prompt but created a year apart reveal noticeable differences.
Get a Glimpse of ChatGPT's Transformation
Ever wondered how much ChatGPT has improved in just a year? Let me illustrate it for you, using two images generated from the same prompt but created a year apart.
The prompt I employed was, "Draw a bird."
Looking at these images, you'll notice a stark contrast between the two. A year ago, ChatGPT was still creating images using simple geometric shapes such as triangles and circles:
Fast-forward to today, and when I type the same prompt, I receive this:
The evolution is astounding, isn't it? It's hard to believe that just a year separates these two masterpieces.
Soaring High with Improvements
Recently, I moved up to an Enterprise account, and during the switch, ChatGPT converted all my previous conversations to the new account. I stumbled upon my first interaction with ChatGPT from almost a year ago, when I first took up writing about AI on this platform.
One of the first things I asked ChatGPT to accomplish back in 2024 was to draw a bird. This isn't exactly representative of the state of AI image generation a year ago, though. At the time, other tools like Midjourney were producing photorealistic images, but my free ChatGPT could only manage these simple geometric shapes.
If I had ChatGPT Plus, things would have been different. In 2024, I could have asked ChatGPT Plus to use its built-in DALL-E 3, a powerful AI image generator, to produce a photorealistic image of a bird. It wouldn't have been as good as the current results, but it would have looked significantly better.
DALL-E 3 access was granted to all ChatGPT Plus accounts in October 2023, but it wasn't integrated into the free version of ChatGPT in the same way. It wasn't until August 2024 that free ChatGPT users could start generating images using DALL-E, but were limited to just two image requests a day. Once the limit was reached, users were back to creating basic-looking images like the one depicted above.
Pushing Boundaries
These days, it's almost unbelievable that ChatGPT ever produced such simple images. Since ChatGPT introduced its new native image generation engine in March this year, we've grown accustomed to AI image generation of high quality from platforms like ChatGPT and Google Gemini.
In fact, it's rare that a week goes by without a new trend involving ChatGPT-inspired images sweeping social media. It all began with the Studio Ghibli-inspired images and progressed to include the action figure trend, and more.
Reflecting on how far we've come in just a year is a truly remarkable experience. With each passing day, I become increasingly curious about what the future holds.
Will anything less than a fully realistic image with automatic video generation, including sound, be construed as a failure in 2026? It's possible, or maybe we'll all be using AI glasses by then, allowing us to step inside our images and explore them from different angles.
Whatever comes next, I can't wait to find out.
Bonus content
- How does ChatGPT actually create its responses? Discover the secrets behind its AI
- From screen to silver screen: This blockbuster movie features an entirely AI-generated cast and production - learn 3 key things about it
- Unleash the power of ChatGPT: Reddit users share 10 unexpected ways to use ChatGPT like a pro
Extra enrichment:
Currently, OpenAI hasn't disclosed a specific timeline for integrating fully realistic video generation with sound into the free version of ChatGPT. However, recent developments include:
- AI Image Generation: OpenAI introduced a more advanced AI image generation model in March 2025, enabling users to produce detailed images directly on the platform.
- Image Library: In April 2025, ChatGPT introduced an image library for organizing and managing AI-generated images.
- GPT-4 Replacement: As of April 30, 2025, the GPT-4 model was phased out, and GPT-4o became the default model, which offers improved capabilities in various areas.
- Memory Improvements for Free Users: In early June 2025, OpenAI rolled out memory improvements for free users, enhancing their overall experience with ChatGPT.
As for the introduction of video generation with sound, no specific timeline has been announced for the free version of ChatGPT. However, the continuous advancements in AI technology by OpenAI may lead to the exploration of such features in future updates.
- The evolution in AI image generation is remarkable, with ChatGPT's new native image generation engine producing high-quality images that contrast starkly with the simple geometric shapes it was creating just a year ago.
- Today, we've grown accustomed to platforms like ChatGPT and Google Gemini offering AI image generation of high quality. It's fascinating to consider the future advancements in AI technology, such as fully realistic image and video generation with sound, which may become the new standard in a few years.
