A Four-Part Prompt Pattern for FLUX 3 Image
Priya N. Chen on 12th Nov 2026
A Four-Part Prompt Pattern for FLUX 3 Image
Black Forest Labs’ FLUX 3 Image turns written prompts into pictures and can edit existing images. Artists can guide a result with four prompt parts: subject, material or medium, camera or mix, and constraint. A prompt is the written instruction sent to the model, the image-making system that produces an image. The tool can also combine reference images—pictures supplied to guide a result—and place elements with layout controls.
Start with the subject and setting. The maker’s example is a risograph illustration of a rabbit riding a pumpkin. Risograph is a print method whose look can be imitated in an image. Next, name a material or medium, such as that print-like style. Then describe the light and framing: how the scene is lit and arranged in the picture. These details steer different parts of the result. The subject says what to show; the medium suggests its visual treatment; light and framing guide its appearance. For the fourth part, give a clear constraint or, when editing, state exactly what to change. The overview says prompts can describe the subject, setting, light, and framing, and recommends stating exactly what to change when editing.
For layout, FLUX 3 Image uses bounding boxes—coordinates that specify where elements should be placed. Their coordinates follow the order [top, left, bottom, right] and use integers from 0 to 1000. The system can combine up to ten reference images. The hope is faster visual exploration and more room to test ideas. The fear is that AI image tools could reduce commissions and concentrate creative power. Geoffrey Hinton has warned that machines may out-think us sooner than expected; Yann LeCun argues against doomerism and says current systems are far from human-level. Is AI a tool in the hand of the artist, or something else? In this case, the artist directs the prompt and decides what to keep, while the system generates or edits the image.
To start, a user creates an account, adds credits, makes an API key—a private code used to send requests—and makes an API call, or request to the system. Black Forest Labs suggests starting with $10–20 to experiment. The maker says pricing is per image and varies by output resolution; one credit equals US$0.01. The system handles requests asynchronously, meaning it returns a polling link that the user checks for the finished image. The artist supplies the direction and decides what to keep. The model supplies an image that may need further edits. That makes the prompt and the artist’s choices central to the work.
Join Neoclassical Pop Art daily take on the arts, the weekly letter, and first word on product launches, sales, and free gifts.