ChatGPT Images gets GPT‑Image‑1.5: faster, more precise, and easier to create

Read Articleadded Dec 17, 2025
ChatGPT Images gets GPT‑Image‑1.5: faster, more precise, and easier to create

OpenAI released a new ChatGPT Images experience powered by GPT‑Image‑1.5, now available in ChatGPT and the API. It delivers more precise edits, stronger instruction following, better text rendering, and up to 4× faster generation, plus a new sidebar creation space with presets and likeness reuse. The API adds improved brand/logo preservation and 20% lower image I/O costs, with OpenAI noting clear gains alongside remaining limitations.

Key Points

  • New model and experience: ChatGPT Images now runs on GPT‑Image‑1.5, rolling out in ChatGPT for all users and in the API.
  • Precision and preservation: Stronger instruction following and precise edits maintain lighting, composition, and likeness across iterative changes.
  • Creative power and text: Expanded creative transformations and improved dense/small text rendering produce more natural, faithful results.
  • Speed and workflow: Image generation is up to 4× faster with concurrent creation, plus a new sidebar creation space with presets and likeness upload.
  • API upgrades and cost: GPT Image 1.5 offers better brand/logo consistency and is 20% cheaper per image I/O than GPT Image 1, supporting marketing and ecommerce use cases.

Sentiment

The overall sentiment of the Hacker News discussion regarding GPT Image 1.5 is decidedly mixed, leaning towards cautious skepticism and criticism despite acknowledging specific technical improvements. While some users praise its advancements in prompt understanding, localized editing, and specific creative use cases like 'previz-to-render,' a significant portion expresses disappointment with its image fidelity, inconsistent prompt adherence compared to leading competitors, and ethical concerns surrounding content authenticity and copyright. There's also frustration directed at OpenAI's rollout execution and perceived lack of diversity in examples.

In Agreement

  • GPT Image 1.5 significantly improves localized edits without altering overall image aesthetics, doubling previous benchmark scores and passing challenging prompts like 'Giraffe'.
  • The model shows strong prompt understanding and steerability, with high compliance rates (e.g., 90% in GenAI Showdown).
  • It excels at 'previz-to-render' tasks, understanding scene layout and character poses from low-fidelity inputs and upscaling them, outperforming Nano Banana (Pro) in this specific area.
  • GPT Image 1.5 can replicate intricate details and follow complex instructions for specific image modifications (e.g., 'Lord of War' mosaic, generating specific texture maps).
  • The previous 'yellow cast' or 'piss filter' issue in OpenAI's image generation appears to be resolved in this new version.
  • The technology is seen as 'magical' and a 'major win for the world' for democratizing visual expression and articulating ideas visually, similar to the early impact of computing.

Opposed

  • Despite improvements, many users find GPT Image 1.5's overall image fidelity, artistic quality, and prompt adherence to be inferior to competitors like Nano Banana Pro and Midjourney, especially for stylistic work or complex spatial reasoning.
  • There are significant concerns about the immediate availability and reliability of the API, with users reporting 500 or 400 errors and criticizing OpenAI's communication regarding 'staggered rollouts'.
  • Ethical issues like the 'death of copyright', the use of personal appearance data, potential misuse for generating images of children, and the erosion of trust in genuine content are frequently raised.
  • The energy consumption of AI image generation is seen as excessive, particularly for perceived 'slop' or commercial ads, leading some to question the sustainability and overall value proposition.
  • OpenAI's own marketing examples are criticized for a significant lack of diversity, featuring an overwhelming majority of white individuals.
  • Some users perceive the cost-to-value ratio as poor, expressing disillusionment with OpenAI for charging more for 'shitty images'.
  • The effectiveness of watermarking and other methods to distinguish AI-generated content from real content is doubted, leading to fears of an intractable problem regarding media authenticity.