Adobe Unveils Next-Gen Firefly Image Models with an All-New Web Experience
Adobe recently launched new AI models in its Firefly series. These include a new and improved model for creating images, a model for creating vector art, and a new web app that offers AI tools from Adobe as well as other companies. Along with this, a mobile app is also going to be launched.
According to Adobe, the new Firefly Image Model 4 offers better quality, speed, and control than before. In this, users can control things like style, angle and zoom of photos well. This model can create images in quality up to 2K. Image Model 4 Ultra is a version that is capable of creating complex and intricate scenes with extreme precision and micro details.
Adobe’s VP of Generative AI, Alexandru Costin, said that this time the model has been trained with more powerful computing, due to which it is now able to create images in more detail. Along with this, now this model also gives better results in image generation with text and users can also apply the style of their favorite pictures.
Adobe has also made its Firefly video model, which was previously in limited testing, available to the public. This model can create video clips from text or images. Users can control elements such as camera angle, start and end frames, movement design and environment. This model provides video quality up to 1080p.
Another new model of Firefly — Vector Model — now allows users to create vector-based artwork such as logos, icons, packaging designs and patterns, which can also be edited later.
Firefly’s new web app brings all these models together, as well as gives the option to access AI models from companies like Google, OpenAI and Flux. Users can create images from any model and content information (such as who created it, how it was made) is attached to them. The company also said that more models can also be added to this platform in the future.
image credit : ADOVE
Adobe is also testing a new tool called “Firefly Boards”, where users can edit, remix and collaborate on creative ideas with others. This tool is compatible with platforms like Visual Electric, Cove, and Kosmik.
Related this :- Which AI Tools Can Create Images in 2025 ?
The company also said that these new AI models will soon be included in the rest of Adobe’s products, but no fixed date has been given for this at the moment.
Along with this, Adobe has opened its Text-to-Image API and Avatar API to everyone, while the Text-to-Video API is currently available in testing mode. All these services will be available through Adobe’s Firefly Services platform.
Adobe is also testing another web tool called Adobe Content Authenticity. Through this, users can add ownership information to the content they create, and can also decide whether their images should be used for AI model training or not.
