# OpenAI Releases ChatGPT Images 2.5 with 50% Latency Cut and Sketch Mode > The September 8 launch adds precise editing tools and faster performance while GPT-6 Astra expands from limited organizations to ChatGPT Plus, Pro, Business and Enterprise users plus the API. *Published 2026-09-10 · By The Intel Desk* ChatGPT Images 2.5 is OpenAI's state-of-the-art image model featuring sharper details, more precise editing, and faster generation. OpenAI launched ChatGPT Images 2.5 on September 8, 2026, as its new state-of-the-art image model featuring sharper details, more precise editing, and faster generation. The update responds to high demand for image tools that support rapid iteration and reliable results across creative workflows. ## What context led to the ChatGPT Images 2.5 and GPT-6 Astra releases? Image generation volumes have grown substantially, with OpenAI reporting that users create more than 3 billion images per week across ChatGPT Images and GPT-Image API models. This scale highlights the need for continued improvements in speed and control to support everyday use cases in design, content creation, and professional applications. The company has iterated on prior image models to address latency and editing accuracy. GPT-6 Astra, described as OpenAI's most capable model with advances in computer use, coding, and professional workflows, began its rollout on September 3, 2026, to a limited set of organizations before expanding further. ## What new features does ChatGPT Images 2.5 introduce? Sketch mode lets users draw references directly in ChatGPT by typing @Sketch, providing an intuitive method to guide generation with custom inputs. Templates offer structured starting points for formats such as Poster and Merch, reducing setup time for common project types. Inline comments on generated images allow targeted instructions for specific areas, improving the precision of multi-turn edits. Users can also share prompts alongside images to enable collaboration and prompt reuse across teams or projects. The model is available to all ChatGPT, ChatGPT Work, and Codex users on desktop, mobile, and web. This broad distribution ensures consistent access for individual creators as well as enterprise teams relying on integrated AI tools. ## How does the latency reduction affect image workflows? ChatGPT Images 2.5 reduces image generation latency by up to 50% compared with Images 2.0. Faster responses support more fluid creative sessions where multiple versions must be produced and reviewed quickly. ## What distinguishes GPT-Image-2.5 Flare from GPT-Image-2.5 Sunburst in the API? For API users, OpenAI introduced GPT-Image-2.5 Flare as the default model that delivers higher quality at 50% lower latency than GPT-Image-2. GPT-Image-2.5 Sunburst targets premium precision workflows where longer generation times are acceptable in return for additional detail. Comparison of GPT-Image-2.5 API VariantsModelLatency ProfilePrimary Use CaseGPT-Image-2.5 Flare50% lower than GPT-Image-2Default higher quality workflowsGPT-Image-2.5 SunburstLonger generation timesPremium precision workflows ## What is the rollout timeline for GPT-6 Astra? GPT-6 Astra began rolling out on September 3, 2026, to a limited set of organizations. Over the following days the model became available to ChatGPT Plus, Pro, Business, and Enterprise users as well as through Codex and the OpenAI API. - Limited initial release to select organizations on September 3, 2026. - Expansion to ChatGPT Plus, Pro, Business, and Enterprise users. - Further availability through Codex and the OpenAI API in subsequent days. ## What market and stakeholder implications follow from these updates? Broad availability to ChatGPT Work and Codex users positions the tools for professional adoption in marketing, design, and content pipelines. The combination of reduced latency and improved editing precision lowers barriers for teams producing high volumes of visual assets. C2PA metadata and invisible watermarking on both image models and GPT-6 Astra provide built-in provenance tracking, which may influence how organizations handle generated content in regulated or collaborative environments. ## What expert reactions address the editing capabilities? Industry observers have noted the reliability of multi-turn editing and the ability to preserve visual identity during changes. > Images 2.5 produces more natural lighting and richer textures, is better at preserving the subjects in your reference photos, and follows editing instructions more reliably across multiple turns.OpenAI > What impressed us most about GPT‑Image‑2.5 Flare is how well it understands what not to change. You can make a meaningful edit without losing the character, composition or visual identity of the original image.Axultan Alimkulov, Head of Product at Higgsfield AI ## What can users expect next from OpenAI? Both ChatGPT Images 2.5 and GPT-6 Astra incorporate C2PA metadata and invisible watermarking to support content provenance. Ongoing platform integrations suggest further refinements to creative and professional workflows. The updates to Codex alongside Astra indicate continued development of harnesses that combine image generation with advanced model capabilities across OpenAI's product ecosystem. ## Sources 1. [ChatGPT Images 2.5 reduces image generation latency by up to 50% compared to Images 2.0 and introduces Sketch mode and templates.](https://openai.com/index/introducing-chatgpt-images-2-5/) 2. [GPT-6 Astra began rolling out on September 3, 2026, to limited organizations and will broaden to ChatGPT Plus, Pro, Business, and Enterprise users as well as the API.](https://openai.com/index/gpt-6-astra/) 3. [ChatGPT Images 2.5 is faster, sharper, smarter with Sketch and templates rolling out to all ChatGPT, ChatGPT Work, and Codex users.](https://x.com/OpenAI/status/2097394956457623964) --- Source: https://aiintelreport.com/frontier-models/openai-chatgpt-images-2-5-gpt-6-astra-rollout Index: https://aiintelreport.com/llms.txt · Full text: https://aiintelreport.com/llms-full.txt