ADOBE
Adobe has launched the newest version of its Firefly AI image tools, alongside a redesigned web app that brings together all its generative models, plus a few from competitors.
A mobile version is also on the way.
At the centre of it all is Firefly Image Model 4, which now offers sharper results, faster speeds, and more control over the final look, think camera angles, zoom, and style tweaks.
There’s also Image Model 4 Ultra, a more powerful version that handles complex, detailed scenes with ease.
The video model, which had been in limited beta, is now open to everyone.
It can turn a text prompt or image into a 1080p video clip, complete with motion design, start and end frame control, and atmospheric elements.
Adobe also introduced a new vector model that’s handy for generating logos, icons, and product designs in editable formats.
Other updates worth noting:
Firefly Boards, a collaborative canvas for remixing, moodboarding, and co-creating visuals.
New Text-to-Image and Avatar APIs are now generally available.
A Text-to-Video API has launched in beta.
Adobe is testing Content Authenticity, a tool to tag digital work with ownership info and opt out of model training.
Tools on Tools on Tools
All of this lives inside the refreshed Firefly web app, where users can also switch between AI models from Adobe, OpenAI, Google, and Flux.
Each generated piece includes built-in metadata for transparency and attribution.
Adobe didn’t just pull up, they brought the entire toolbox.