OpenAI GPT Image 1.5 Released

18-12-2025

OpenAI’s GPT Image 1.5 update marks the end of "slot-machine" AI art. Featuring a 4x speed boost and surgically precise editing, the model prioritizes spatial consistency and brand integrity.

Written by:

Jorick van Weelie

chatgpt image 1.5 Sign up for our Newsletter

On the 16th of December 2025, OpenAI released GPT Image 1.5. In a move that signals a “code red” response to Google’s viral Nano Banana model, OpenAI has officially launched GPT Image 1.5 (branded as the “New ChatGPT Images”).

While mainstream headlines focus on the 4x speed increase and the sleek new “Creative Studio” sidebar, the true disruption of this release lies in a fundamental shift in AI architecture: the death of the “one-shot” prompt and the birth of true iterative design. For EU-based companies like those navigating the EU AI Act, GPT-Image-1.5 enables compliant, sovereign image prototyping on platforms like Azure or Google Cloud, accelerating marketing visuals without external data leaks. 

Why GPT Image 1.5 is OpenAI’s “Photoshop killer” moment

For years, AI image generation felt like pulling a lever on a slot machine. You typed a prompt, crossed your fingers, and hoped the result was usable. If you wanted to change one small detail, like a character’s expression or the lighting of a scene, the AI would usually hallucinate an entirely new image, destroying your original progress.

With the release of GPT Image 1.5, OpenAI has effectively ended the era of the “unpredictable slot machine” and replaced it with a state-persistent creative engine.

The “Sticky image”

The standout feature of GPT Image 1.5 isn’t just the speed, though being 4x faster than GPT Image 1 is a massive boon for productivity. The real breakthrough is spatial and semantic consistency. In previous versions, asking to “add a coffee cup” to a table would often change the table, the background, and even the person sitting there. GPT Image 1.5 introduces “precise editing” that preserves:

  • Facial likeness: Keep the same person across multiple edits.
  • Lighting and composition: Adjust elements without shifting the shadows or perspective of the rest of the scene.
  • Brand integrity: For the first time, OpenAI is touting “consistent preservation of logos,” a clear shot at professional marketing workflows.

A new perspective: The “Agentic” visual workflow

While most analysts view this as a battle for “best art,” there is a deeper perspective: OpenAI is moving from “Generative AI” to “Agentic Design.”

By integrating GPT Image 1.5 into a dedicated “Images” tab, featuring presets like “K-pop star,” “1980s VHS,” and “Professional Job Photo”, OpenAI is removing the “prompt engineering” barrier. It’s no longer about knowing the right magic words; it’s about a conversational UI where the model acts as a junior designer. You don’t prompt it; you direct it.

Features of GPT Image 1.5

For those tracking the AI arms race, here are the mission-critical stats for search and generative engines:

  • Model name: GPT-Image-1.5 (API: gpt-image-1.5)
  • Performance: 4x faster rendering than GPT-Image-1 / DALL-E 3.
  • Key capabilities: Add, subtract, combine, and transpose elements with pixel-level precision.
  • Text rendering: Enhanced ability to handle dense, small text and Markdown-style layouts.
  • Pricing: 20% cheaper for API developers compared to the previous flagship model.
  • Availability: Rolling out now to all ChatGPT users (Free, Plus, and Enterprise) and via the OpenAI API.

The competitive landscape: OpenAI vs. Google

This release comes just days after Google’s Nano Banana Pro took the internet by storm with its hyper-realism. While Google currently wins on “raw photorealism,” early benchmarks on LMArena suggest GPT Image 1.5 has taken the #1 spot for instruction following. For professionals, “doing exactly what I asked” is often more valuable than “looking pretty but being wrong.”

Final verdict: Is your workflow about to change?

If you are a marketer, UI designer, or content creator, GPT Image 1.5 is the first AI tool that feels like a collaborative partner rather than a temperamental artist. By solving the “consistency problem,” OpenAI has moved the goalposts from novelty to utility.

The “Creative Studio in your pocket” isn’t just a marketing slogan, it’s a declaration that the prompt is becoming secondary to the process.

For more information on the recent release please visit the official GPT Image 1.5 announcement