AI image generator with FLUX: text-to-image on RafayGen's own GPU

Ask RafayGen for "a watercolor painting of Lahore's old city at sunset" and it paints one. Image generation on RafayGen runs on FLUX — a state-of-the-art open image model — executed on RafayGen's own GPU hardware rather than a rate-limited third-party image API.

What the image studio can do

• Text-to-image: describe a scene, get an image in the chat.
• Image editing: upload a picture and describe the change — swap the background, change a colour, remove an object.
• Img2img: reimagine an existing image in a new style while keeping its structure.
• Upscaling: enlarge an image while sharpening detail instead of blurring it.
• Describe: give it an image and get a detailed description — useful for alt text and cataloguing.

Why running our own GPU matters

Because generation happens on RafayGen's own hardware, images aren't watermarked by a third-party free tier, prompts aren't shaped by someone else's content pipeline, and the studio can expose the full workflow — editing, upscaling and multi-step "studio planner" jobs — instead of a single generate button.

How to use it

In the chat, just ask for an image; the assistant detects an image request and routes it to the GPU studio. You can iterate in replies: "same scene but at night", "make it 4k", "remove the car on the left".