OpenAI Integrates Image Models into Adobe's Creative Studio
OpenAI now supplies the image models inside Adobe’s creative studio. It is also turning prompts into shareable objects.
ChatGPT Images 2.5: Key Features
ChatGPT Images 2.5 ships with two API models, one of which is already integrated into Adobe’s Firefly. More than 3 billion images are created weekly across ChatGPT and the image API.
- Sharp detail, precise editing, and reduced latency: The model offers these improvements over its predecessor, Images 2.0.
- Two new models for developers: GPT-Image-2.5 Flare (default) and Sunburst (slower, higher-control option for production creative work).
Adobe Integrates ChatGPT
Adobe is a launch customer, with the new GPT-Image-2.5 models now available in their creative AI studio, Firefly. This strategic move positions Adobe as providing an interface for various AI models, including ChatGPT’s.
The Competitive Landscape
The integration highlights the competitive question in creative software: who owns the starting point of a creative task. Both companies have established distribution channels for each other, making it difficult to predict which one will dominate.
Social Features in ChatGPT
OpenAI introduces social features, such as:
- Sharing prompts with images: Users can attach the prompt that generated an image, allowing others to run it with their own photos.
- Sketching and templates: Users can draw within ChatGPT, use prompts as references, and access starting formats like posters and merchandise.
Availability and Scale
Images 2.5 is rolling out to all ChatGPT, ChatGPT Work, and Codex users across all tiers, on desktop, mobile, and web.
OpenAI states that more than 3 billion images are created weekly across ChatGPT Images and GPT-Image models, with C2PA metadata, invisible watermarking, and prompt checks in place to ensure provenance.