Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Adobe hasn’t shared a specific date yet, but Firefly Image Model 5 will launch “in the coming months” after Adobe MAX. The Firefly Image Model 4 launched in April 2025, so it will likely be 2026 before we see a full-scale release of the Image Model 5. Like the Image Model 4, the updated model has a native resolution of 4 megapixels, which means it can create images at 2K (2560 x 1440) resolution. It also supports instant editing, which sees edits created at 2MP or Full HD (1920 x 1080). However, the big improvement in Image Model 5 is multi-layer image editing.
In the demo, Adobe showed me how the new Firefly template works. You can upload an image, and Firefly Image Model will select 5 different elements, allowing you to move, resize, and replace those elements with generative features. In my demo, Adobe used a bowl of ramen, demonstrating how Image Model 5 was able to cut chopsticks and move them to a different area of the scene, as well as adding an AI-generated bowl of chili flakes. And all this was done without any visual effects.
New generative features in Adobe Firefly
Firefly is also getting two new AI features: soundtrack generation and speech generation. They both do exactly as their names indicate, with some specific guardrails.
Create an audio clip will scan an uploaded video and suggest a prompt for an audio clip. Instead of deleting the prompt and starting from scratch, Adobe lets you choose the look, feel, and purpose to find something that works. For example, you could say that you want a tense orchestral piece of music to be layered over a chase scene.
Courtesy of Adobe