Search results for GPT Image 2
AI & Enterprise
ChatGPT Image 2.5 vs Nano Banana 2: AI image generators face off - who wins?
OpenAI’s new image model, ChatGPT Image 2.5, has improved an over-sharpening issue flagged in its previous version, but a comparison with Google-side Nano Banana 2 was too close to call. In six evaluation categories, the two models each won three, producing a near draw. ChatGPT Image 2.5 performed better in spatial composition, mood and illustrations, while Nano Banana 2 more reliably handled verifiable details such as text and dates.
AI & Enterprise
Alibaba unveils Qwen Image 3 for generating full newspaper pages
Alibaba has unveiled Qwen Image 3.0, a generative AI image model designed to create complex work products such as newspaper pages, infographics and storyboards in a single generation. The company said the model can process prompts of up to 4,500 tokens and render small text and LaTeX math notation, while supporting 12 languages and simulating digital interfaces. It did not release open weights, benchmarks, technical reports or API pricing.
AI & Enterprise
ChatGPT, Gemini Omni can make ad videos for about 1,500 won in computing costs
A single product photo and several generative AI tools can produce about 10-second product ads for TikTok and YouTube at roughly $1 in computing costs, the article says. The process uses ChatGPT for a virtual model image and a short script, Google Flow with Gemini Omni to generate video with audio, and CapCut for final edits. Platforms require labeling AI-generated content and disclosures for commercial promotions.
