Adobe Inc., a leading creative software company, is enhancing its product lineup with additional artificial intelligence models tailored to meet the evolving needs of professionals in their daily workflows.
The company unveiled Firefly Boards, an AI-native workspace designed for real-time exploration, iteration, and collaboration. This platform will integrate new AI-generated video models, including Aleph from Runway AI Inc. and Marey from Moonvalley AI.
Adobe Photoshop, the company’s flagship image editor, will now support Gemini 2.5 Flash Image from Google LLC—popularly known as “Nano Banana”—as well as FLUX.1 Kontext Pro from Black Forest Labs Inc.
Adobe is rolling out Firefly Boards globally, accompanied by new features such as presets, generative text editing, and image description tools.
The preset feature enables users to generate images in various styles. For instance, the “Products and Characters” preset helps visualize products through image generation, while the “Virtual Try-On” feature allows users to digitally dress models in different outfits.
Currently in beta, the generative text editing function allows users to replace text directly within images while maintaining the original style and context. For example, a user can change the text on a neon sign or billboard and preserve the look of the original neon letters, including font, structure, color, and orientation.
“With Firefly Boards, I can rapidly generate and iterate on multiple visual concepts for creative business proposals,” said Malick Lombion, filmmaker and creative director.
The new image description feature allows users to generate natural language prompts that can be used for image creation. These prompts can later be edited to suit specific needs. For example, a user can upload an image of a colorful bird and convert it into a text prompt, which they can then modify to change the environment, colors, species, or other characteristics before rendering a new image using Adobe’s image generation model.
Photoshop’s Generative Fill Feature Gets a Major Upgrade
Photoshop already integrates Adobe’s Firefly AI models, which are commercially safe for use in creative workflows. The addition of Gemini 2.5 Flash Image and FLUX.1 Kontext expands the creative possibilities within the Generative Fill feature, allowing users to add, remove, or transform parts of an image using natural language prompts.
“This marks the first integration of partner models into Creative Cloud applications,” said Joel Baer, Director of Product Management at Adobe, in an interview with SiliconANGLE. “This is just the beginning of our journey to incorporate these models into Photoshop.”
Gemini 2.5 Flash Image excels at generating stylized elements, graphic details, and scene additions while maintaining contextual awareness. It also delivers more consistent results during fine edits, avoiding unintended alterations.
FLUX.1 Kontext is optimized for contextual accuracy, perspective, and environmental harmony. It performs exceptionally well in complex tasks such as adjusting settings, handling human anatomy, or changing backgrounds while preserving the core elements of the scene.
When using these models, Photoshop automatically generates non-destructive layers that record the prompts and models used. This enables creatives to refine their edits, apply masks, and keep AI-generated changes separate from the main image.
For example, designers can use Flash Image to add props or transform a wine glass into a champagne flute, while FLUX.1 can change a wooden desk to glass without affecting the items placed on top.
“Many of our customers already use third-party services alongside Photoshop, whether for ideation or image editing,” Baer said. He added that integrating these new models will expand creative workflow options. “Photoshop has always been the go-to when you're unsure of what you need to do to an image—but you know Photoshop can handle it.”
Adobe announced that during the testing phase, the new Image Flash and FLUX.1 models will be temporarily available to all subscribers at no cost and without deducting AI credits. Standard plan subscribers will receive 100 lifetime generations per model, while existing premium subscribers will get 500 daily generations per model.
