Adobe has significantly expanded its Firefly suite of generative AI tools, introducing two new image generation models—Firefly Image Model 4 and Firefly Image Model 4 Ultra—and integrating third-party AI models from OpenAI and Google into its platform. These developments enhance creative possibilities while maintaining Adobe’s commitment to commercial safety.
🖼️ New Firefly Image Models: 4 & 4 Ultra
Adobe’s latest models offer enhanced capabilities for creators:
- Firefly Image Model 4: Prioritizes speed and efficiency, enabling users to generate images up to 2K resolution with improved control over style, format, and camera angles.
- Firefly Image Model 4 Ultra: Designed for generating more complex and detailed scenes, providing higher fidelity and creative freedom.
These models are now accessible via the Firefly web app, which also introduces Firefly Boards, a collaborative moodboarding tool currently in public beta. Additionally, Adobe plans to launch a mobile Firefly app for iOS and Android soon
🤖 Integration of Third-Party AI Models
Expanding creative options, Adobe has integrated several third-party AI models into Firefly:
- OpenAI’s GPT Image Generation
- Google’s Imagen 3, Veo 2, and Flux 1.1 Pro
These integrations allow users to ideate using various models within Firefly and seamlessly transfer generated content into Adobe applications like Photoshop. All models, including third-party ones, can be accessed using the same Adobe credit system. However, only Adobe’s own models are deemed “commercially safe” due to their training on licensed or public content.
🎨 Enhanced Features in Adobe Creative Cloud
Adobe continues to enhance its Creative Cloud applications with AI-driven features
- Photoshop: Introduces advanced tools like Generative Fill, Generative Expand, Generate Similar, Generate Background, and Enhance Detail, all powered by the Firefly Image 3 Model. These tools offer improved photorealistic quality, better understanding of complex prompts, and greater variety in generated outputs.
- Illustrator: Implements features like Objects on Path for arranging objects along any path and enhancements to the Image Trace tool for crisper vector outputs.
- Premiere Pro: Introduces the Generative Extend feature, currently in beta, powered by the Firefly AI Video Model, allowing for seamless video editing and extension.
🔐 Commitment to Commercial Safety
Adobe emphasizes that its Firefly models are trained on licensed content, such as Adobe Stock, and public domain content where copyright has expired, ensuring that the generated images are safe for commercial use. This approach addresses concerns surrounding generative AI and its impact on content creators, promoting trust and transparency in digital content.