A new era of image manipulation has arrived within ChatGPT, powered by the latest update, GPT Image 1.5. This isn’t a simple upgrade; it’s a fundamental shift in how we interact with and alter visual content, rolling out to all users now.
One of the most significant advancements mirrors a feature recently introduced by Google Gemini: the ability to precisely edit specific areas of an image while preserving the integrity of the rest. Imagine seamlessly adding or removing elements, or altering colors and styles, without the telltale signs of artificial modification.
ChatGPT now effortlessly combines multiple images into a single, cohesive scene. Want a portrait with a friend in front of a famous landmark? Simply provide the source photos, and the AI intelligently merges them, creating a realistic composite.
OpenAI reports that GPT Image 1.5 follows instructions with greater accuracy and renders images up to four times faster than its predecessor. Text within images is now more versatile, and the overall quality boasts increased realism and fewer errors – though refinement continues.
To truly understand the power of this update, a direct comparison with Gemini was essential. Both platforms were tested using a $20-per-month subscription tier, pushing their capabilities to the limit with a series of challenging prompts.
The first test involved generating complex scenes: a busy tech journalist, a solitary lamp in a vast warehouse, and a stylized, foggy landscape. Both ChatGPT and Gemini delivered impressive results, with quality and realism remarkably similar, though occasional inconsistencies appeared.
Both AI models excelled at clean image edits, flawlessly changing a journalist’s attire without affecting surrounding details. This level of precision, once requiring expert-level Photoshop skills and significant time investment, is now achievable in seconds.
However, perspective shifts proved more challenging. While both AIs attempted to render the same shot from a different angle, the results were less consistent, with ChatGPT demonstrating a slight edge in maintaining image coherence.
The classic “remove an object” test showcased remarkable precision. Both Gemini and ChatGPT surgically removed a cottage from a countryside scene, leaving the surrounding environment undisturbed – a task previously demanding meticulous manual effort.
Combining images presented a new set of hurdles. While both platforms could merge photos, achieving a truly seamless and believable result required careful prompting and tweaking. Lighting inconsistencies and scaling issues were common challenges.
ChatGPT consistently outperformed Gemini in blending images and altering overall aesthetics. When tasked with creating a moody film noir scene from multiple sources, ChatGPT produced a cohesive image, while Gemini’s attempt resembled a simple cut-and-paste.
Remixing photos – adding people, changing weather, altering locations – is now incredibly accessible. However, generating realistic depictions of known individuals remains a challenge, as the AI lacks intimate knowledge of their unique features and mannerisms.
Currently, ChatGPT holds a slight advantage over Gemini in image editing capabilities, but the gap is narrow. Both platforms have democratized advanced image manipulation, placing powerful tools within reach of everyone.
The future of AI-powered image editing is undeniably bright. These advancements are not merely about convenience; they represent a fundamental shift in how we create, modify, and experience visual content.
