Adobe Firefly, the company’s suite of generative AI models, received significant upgrades at Adobe MAX London 2025. It introduces the powerful new capabilities in image and video generation. It also enhances photo editing and vector graphics. This reinforces its role in AI-assisted creativity.
The focus is on enhanced quality, greater user control, and expanded integration, including options for using partner AI models.
Key takeaways:
- Enhanced models & quality: Introduction of Firefly Image Model 4/4 Ultra for more realistic images (up to 2K). And the generally available Firefly Video Model for production-ready video generation.
- New creative tools: Firefly Boards for moodboarding, Generative Extend in Premiere Pro, and Distraction Removal in Lightroom.
- Increased flexibility: Firefly now serves as an all-in-one app, integrating partner AI models, offering creators more choice.
Adobe Firefly: Next-generation image and video
Adobe unveiled its advanced image model, Firefly Image Model 4 (and 4 Ultra). It promises more lifelike image generation with better detail rendering. Users have control over structure, style, and camera angles. Outputs up to 2K resolution enhance post-production flexibility.
The Firefly Video Model is now available. This marks a milestone as a safe, IP-friendly AI. It generates 1080p video clips from text or images. It also offers controls for camera settings and motion, empowering creators to craft specific visuals.
Learn more: Adobe Firefly Overview
New tools for seamless workflows
- Firefly boards (beta): An “AI-first” collaborative space for teams to brainstorm, generate image variations, remix ideas, and organize concepts.
- Generative extend (premiere pro): This tool adds frames to the beginning or end of clips. It can add up to 2 seconds for video and 10 seconds for audio. This smooths transitions without slow motion. It also supports 4K and vertical video.
- Distraction removal (lightroom): The “Generative Remove” tool lets photographers erase unwanted objects—like photobombers and stray hairs. AI then generates a seamless fill to finish the image.
- Generative fill: This Firefly feature allows users to add, remove, or replace objects in images using text prompts.
Learn more: Extend video Length with Generative Extend
How to use Adobe Firefly:
- Explore the Firefly web app: Visit the revamped Firefly web app. Experiment with Image Model 4. You can also access partner models.
- Try Firefly boards: Sign up for the public beta to test the new collaborative moodboarding capabilities. It’s perfect for your next creative project!
- Update creative cloud apps: Make sure your Photoshop, Premiere Pro, and Lightroom are up-to-date. This will give you access to features like Generative Extend and Distraction Removal.
- Experiment with prompts: Use Firefly Image Model 4 to refine prompts for better control over structure, style, and angles.
FAQs on Adobe Firefly
1. What is Adobe Firefly Boards?
It’s a new collaborative, AI-powered moodboarding tool in the Firefly app (currently in beta). You can explore and develop creative ideas using generated assets. You can also use uploaded assets.
2. Can Firefly extend video clips?
Yes, the Generative Extend feature in Premiere Pro can add up to 2 seconds to video clips. It can also add 10 seconds to audio clips using AI.
3. How does Distraction Removal work in Lightroom?
The ‘Generative Remove’ tool uses Firefly AI to remove selected objects. It intelligently fills the space based on the surrounding background.
4. Is Firefly safe for commercial use?
Adobe designs its Firefly models (Image, Video, Vector) to be commercially safe. They train these models on licensed content like Adobe Stock and public domain images. Content Credentials are also integrated.
5. Can I use non-Adobe AI models within Firefly?
Yes, the Firefly app now allows the choice to use models from partners like Google Cloud. You can also choose models from OpenAI. There are more to come.
Further Reading
Have you used the Adobe Firefly for generating creatives? Let us know what you think about its capabilities in the comments!
Read more about Generative AI models here:
- Vertical AI Agents Will Replace SaaS – Experts Warn On Future Of Workflow Automation
- Stanford’s S1 model – How Stanford built low-cost open source rival to OpenAI’s o1
- What is OpenAI o3-mini – technical features, performance benchmarks, and applications
- What is Alibaba’s Qwen2.5-Max – technical features, performance benchmarks, and applications
- ChatGPT vs Gemini 2.5 Pro – Analyzing Reddit And Expert Reviews
- Google Gemini 2.5 Pro Goes Free: How To Upgrade Your AI Powers
- Vibe coding using Google Firebase Studio: 3 Ways it Simplifies Vibe Coding For No-coders
- Check out latest with OpenAI: Review of OpenAI o3 vs o4-mini
Get the latest updates about using AI for daily and workplace productivity:
This blog post is written using resources of Merrative. We are a publishing talent marketplace that helps you create publications and content libraries.
Get in touch if you would like to create a content library like ours. We specialize in the niche of Applied AI, Technology, Machine Learning, or Data Science.
Leave a Reply