Adobe has unveiled Firefly 5, its latest image-generation model, which supports native resolutions up to four megapixels. This upgrade includes improved human rendering, better lighting, and advanced editing capabilities. The company emphasizes that all Firefly models are commercially safe and production-ready, setting them apart from competitors like OpenAI's DALL-E and Google's Imagen. Firefly 5 builds on the features of its predecessor, Firefly 4, with enhanced resolution options and more detailed image generation. Key features of Firefly 5 include prompt-based and layered editing. Users can modify specific elements of an image without the need to regenerate the entire piece. For instance, if a creator wants to swap out an object in their image, they can simply enter a prompt to make that change. This approach streamlines the editing process, allowing for more control over the final product. Layered editing is currently available in preview mode. Adobe is also expanding its third-party model offerings, adding new image models from Topaz and an audio model from ElevenLabs. Firefly Custom Models, which allow brands to create tailored AI models using their assets, are now available to all creators in a private beta. This shift opens up more opportunities for customization and efficiency in content creation. Adobe's AI Foundry program further supports businesses by providing a framework for designing and training models on their own data, contrasting the DIY approach of Custom Models with a more hands-on support system from Adobe.