Adobe Firefly Expands Image and Video Creation
In a blog post, the company introduced several new features, all focused on its Firefly AI Studio. Adobe has now moved its Custom Models into public beta, making the ability to train AI on personal images and styles available to the general public. It was first unveiled in March 2024 for enterprise customers, and it was introduced as a private beta in October 2025.
With this, brands and individuals can upload their images and assets on Firefly, and a custom AI model will analyse them for style, characters, and overall visual aesthetic. It also takes note of details like stroke weight, colour palettes, lighting, and character features. Then, when the user generates images using the custom model on the platform, the output will be aligned to the shared visual style.
“For teams producing high volumes of content, that consistency becomes a competitive advantage. And your models are private by default, so the content you create with them remains entirely yours,” the company said.
Firefly users now have access to more than 30 AI models, including Google’s Nano Banana 2 and Veo 3.1, Runway’s Gen-4.5, Adobe’s Firefly Image Model 5, which is now generally available, as well as the new Kling 2.5 Turbo. Additionally, the company is also adding tools such as Quick Cut, which turns raw footage into a structured first cut within minutes.
Another major announcement was moving Project Moonlight to private beta. First announced at the Adobe Max event, it is a conversational agentic interface that works across Adobe apps and automates various steps of design and project creation. Users can simply describe what they want in a chat, and agents begin working to execute that vision.
