OpenAI Launches ChatGPT Images 2.5 With Faster Generation
OpenAI’s ChatGPT Images and GPT‑Image models now generate more than 3 billion images a week, underscoring how image generation is becoming a mainstream creative and commercial tool. ChatGPT Images 2.5 is designed to move workflows beyond one-shot prompting by preserving subjects from reference photos more faithfully, rendering more natural lighting and richer textures, and maintaining details through successive edits. Better handling of complex layouts, transparent backgrounds and real-world information also targets professional uses such as marketing assets, product imagery and rapid visual prototyping.
OpenAI released ChatGPT Images 2.5 on Sept. 8, 2026, cutting image-generation latency by as much as 50% compared with Images 2.0. The update adds Sketch for drawing references directly in ChatGPT, templates for formats such as posters and product photos, comment-based local edits and prompt sharing. It is rolling out across all tiers of ChatGPT, ChatGPT Work and Codex on desktop, mobile and web. OpenAI also launched GPT‑Image‑2.5 Flare and the slower, higher-precision GPT‑Image‑2.5 Sunburst through its API.
All Coverage
2 original reportsThe Backstory
The history behind this eventOpenAI Launches ChatGPT Images 2.0 With Better Text Rendering and Multi-Image Generation
Following GPT-4o Image Generation 1.0 and 1.5, OpenAI has introduced ChatGPT Images 2.0, moving image generation beyond one-off creations toward workflows that can handle research, reasoning and layout. Improvements in text rendering, character consistency and multilingual capabilities could help the advertising, e-commerce and design industries produce ready-to-use materials directly, though Chinese text may still appear garbled.
OpenAI released ChatGPT Images 2.0 on April 21, 2026, under the API model name gpt-image-2. Its Thinking mode can search the web and generate multiple images in sequence. The model is available across all ChatGPT plans. Through the API, image output costs $30 per 1 million tokens, while text and image inputs cost $5 and $8, respectively.
OpenAI Launches GPT-5.3 Instant to Improve ChatGPT Conversations and Accuracy
GPT-5.3 Instant is OpenAI’s fast model for everyday ChatGPT conversations, succeeding GPT-5.2 Instant. Rather than extending reasoning, it focuses on improvements users encounter most directly: tone, relevance and conversational flow. The update also addresses excessive refusals, lengthy warnings and insufficient context in web-search responses, all of which affect reliability across ChatGPT’s most common use cases.
OpenAI rolled out GPT-5.3 Instant to all ChatGPT users on March 3, 2026, under the API name gpt-5.3-chat-latest. It did not disclose a separate price for the update. Internal evaluations showed hallucination rates in high-stakes domains fell 26.8% with web access and 19.7% without it, while errors reported by users declined 22.5% and 9.6%, respectively.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.
If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →