Did ChatGPT Just Kill Nano Banana?

Did ChatGPT Just Kill Nano Banana?

by Matt Wolfe

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Matt Wolfe 👥 1.0M 📅 December 17, 2025 ⏱ 23 min 👁 328K 📄 news review 🧭 2026-08-28
Available in: English (current) Français

Keywords

GPT Image 1.5Nano Banana Proimage editingAI comparisonOpenAI

Summary

Matt Wolfe reviews OpenAI’s newly released GPT Image 1.5 model, positioning it as a direct competitor to Google’s Nano Banana Pro. He outlines the model’s key features: faster generation (up to 4x), a new dedicated Images tab in ChatGPT with preset styles, improved instruction following, better text rendering, and a buried ’likeness retention’ feature. He then conducts a series of head-to-head tests against Nano Banana Pro, covering image editing, multi-step consistency, spatial reasoning, text rendering, style transfer, crowd rendering, and brand consistency. The results are mixed: GPT Image 1.5 excels at text rendering and multi-step context memory, while Nano Banana Pro often maintains better visual consistency and adheres more strictly to edit instructions. Wolfe concludes that Nano Banana Pro is still slightly ahead overall, but GPT Image 1.5 is catching up quickly, and he notes that this is the worst these models will ever be.

145 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video’s primary value lies in its practical, hands-on comparison of two leading AI image models. Wolfe’s testing methodology is systematic and transparent, covering a range of typical use cases from simple edits to complex multi-step transformations. He provides clear visual evidence for each test, allowing viewers to judge the results for themselves. The argumentation is balanced: he acknowledges strengths and weaknesses of both models, avoiding fanboyism. He also contextualizes the release within the broader competitive landscape of AI development. The main limitation is the subjective nature of the evaluation, as ‘better’ is often a matter of personal preference or specific use case.

Scientific Rigor, Source Quality, Title Accuracy

The video relies primarily on the creator’s own testing and observations, supplemented by information from OpenAI’s official blog post and announcement. The creator is transparent about his lack of early access and his methodology. The title is somewhat sensationalist but accurately reflects the video’s central question. The video includes a sponsorship segment, which is clearly disclosed. The creator’s own website (futuretools.io) is mentioned as a resource for AI tools and news, but it is not a primary source for the technical claims. Overall, the sourcing is adequate for a review video, though it lacks independent verification.

214 words

Title / Content Match

The title is a catchy question that accurately reflects the video's core comparison between ChatGPT's new image model and Nano Banana Pro.

Quality & Reliability

7/10

The video is a hands-on review of a newly released AI image model, based on direct testing and comparison with a competitor. The creator is transparent about his methodology and limitations (no early access), and he provides balanced observations. However, the analysis is subjective and not peer-reviewed, and some claims rely on the model's official announcements.

Chapters

Cited Sources

  • Future Tools — Creator's website for AI tools and news, mentioned as a resource.
  • Future Tools Newsletter — Creator's weekly newsletter, mentioned as a resource.
  • Webflow — Sponsor link, used to build the creator's website.
  • Matt Wolfe on LinkedIn — Creator's social media profile.
  • Matt Wolfe on Threads — Creator's social media profile.

Concurring Sources

  • OpenAI's GPT Image 1.5 announcement — Official source for the model's features and capabilities, which the video references.
  • Nano Banana Pro (Gemini 3) announcement — Official source for the competing model, which the video compares against.

Dissenting Sources

  • User reports of censorship in ChatGPT image generation — Several comments mention that OpenAI's content policies restrict certain prompts, which could be seen as a limitation not fully addressed in the video.

Contribution & Novelties

The video provides a timely, hands-on comparison of two newly released AI image models, offering practical insights for users deciding which tool to adopt. It highlights specific strengths and weaknesses in areas like text rendering, multi-step editing, and consistency, which are not always covered in official announcements. The ’likeness retention’ feature is brought to light, as it was buried in OpenAI’s blog post.

Pour aller plus loin :

  • OpenAI’s GPT Image 1.5 announcement — Official details on the model’s capabilities and improvements.
  • Nano Banana Pro (Gemini 3) announcement — Official details on Google’s competing model.
  • AI image generation — Background on the field and its evolution.

106 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quantity and quality, reflecting the video's comprehensive coverage and practical testing. The lower technical depth score indicates that the content is accessible to a general audience rather than deeply technical.

Reliability 7/10

💬 Positif. Sur les 30 commentaires analysés, le ton est majoritairement enthousiaste et constructif, avec des retours d'expérience et des discussions techniques sur les forces et faiblesses des modèles comparés.