FIBO Mashup is a free AI-powered image combination generator that merges three separate images you provide into a single composite image.
It analyzes the subject image, scene image, and style image you upload, then creates a new image that preserves the main subject from your first upload, places it within the environment from your second image, and applies the artistic style from your third image.
The tool is a practical application of the Bria Visual Language Model (VLM) and its FIBO image model.
The technology is designed for developers and companies to generate and edit visuals, but this free tool on Hugging Face makes a piece of that power accessible to anyone.
It works by having an AI analyze the core components of each picture you provide and then intelligently merge them.
Features
- Three-Image Fusion: Combines subject, scene, and style elements from separate source images into a single generated output.
- VLM-Powered Analysis: Uses Bria’s Visual Language Model to extract detailed structured prompts from each uploaded image before generating the final result.
- JSON-Native Processing: Converts visual inputs into structured data with detailed descriptions of lighting, composition, camera settings, and style attributes.
- Licensed Data Training: Built on 100% commercially licensed training data.
- Free Access: Available through Hugging Face Spaces without requiring API keys or subscription fees.
Use Cases
- Concept visualization: Combine product photos with different environments and artistic treatments for marketing mockups.
- Creative experimentation: Test how subjects appear in various settings with different visual styles.
- Rapid prototyping: Generate composite images for storyboards or project presentations without Photoshop.
- Artistic exploration: Blend photographic elements with painting styles for unique digital art creations.
- Content creation: Produce original images for social media by merging existing assets with new backgrounds and styles.
How To Use It
1. Go to the FIBO Mashup Hugging Face Space.
2. Upload your first image in the Subject Image box. This should be the main person or object you want to feature.
3. Upload your second image in the Scene Image box. This is your desired background.
4. Upload your third image in the Style Image box. This image’s aesthetic will be applied to the final result.
5. Click the ‘🎨 Generate Combined Image’ button.
6. Wait for the process to complete. It can take a minute or two as the AI analyzes the images and the API responds.
7. The final generated image will appear in the “Generated Image” box. You can then right-click to save it.
Pros
- No Cost: It is completely free to use.
- Creative Speed: It automates what would be a multi-step, manual process in photo editing software like Photoshop.
- Unique Results: The combinations can lead to unexpected and highly creative visuals that would be difficult to imagine otherwise.
Cons
- Processing Time: Generation can sometimes be slow, depending on server load and API response times.
- Lack of Control: You cannot fine-tune the result. The AI makes all the decisions about how to blend the images, so what you get is what you get.
Related Resources
- Bria AI Platform: Commercial API access and enterprise licensing options for production use cases that require higher reliability and commercial rights.
- FIBO GitHub Repository: Source code, local installation instructions, and integration examples for developers who want to run the model on their own infrastructure.
FAQs
Q: Can I use FIBO Mashup outputs for commercial projects?
A: The free Hugging Face version is limited to non-commercial use. If you need to use generated images in commercial work (client projects, products, marketing materials), you need a commercial license from Bria AI.
Q: Can I adjust the influence of each input image?
A: No, the current interface doesn’t provide sliders or controls to weight how much each input affects the final result. The AI determines the balance automatically based on its analysis of the three images.
Q: What makes FIBO different from other image mixing tools?
A: FIBO uses JSON-native structured prompts instead of simple text descriptions. When you upload images, the Visual Language Model extracts detailed information about lighting setups, camera angles, composition rules, depth of field, and color relationships. This structured approach provides more precise control and reproducibility than models that rely on simple caption-based training.
