OPEN_SOURCE ↗
HN · HACKER_NEWS// 5h agoMODEL RELEASE
OpenAI drops ChatGPT Images 2.0
OpenAI’s April 21 livestream revealed ChatGPT Images 2.0, a new image generation model with stronger text rendering, aspect-ratio control, multi-image output, and a paid “thinking” mode. It is rolling out to ChatGPT and Codex users, with gpt-image-2 also coming through the API.
// ANALYSIS
This is less about prettier pictures and more about turning ChatGPT into a usable visual production surface for developers, designers, and agents.
- –Text rendering is the real upgrade: menus, posters, UI mockups, maps, and infographics become much more practical when the words stop melting.
- –The Codex angle matters because screenshots, interface concepts, diagrams, and visual specs can now sit closer to coding workflows.
- –API availability means this will quickly show up inside design, marketing, education, and app-builder tools rather than staying a ChatGPT-only feature.
- –The weak spot is still trust: complex factual visuals and non-English text need verification, especially when the model can search and synthesize recent context.
// TAGS
chatgpt-images-2-0openaiimage-genmultimodalapireasoningcodex
DISCOVERED
5h ago
2026-04-21
PUBLISHED
6h ago
2026-04-21
RELEVANCE
9/ 10
AUTHOR
wahnfrieden