Flux.2 Flex

Create text-driven images, marketing assets, and precise typography with FLUX.2 [flex]. Control lettering, layout framing, and visual design.

See what a prompt can become

Find a visual direction, then make the prompt your own.

From a prompt

When to choose it

Layouts where lettering and its placement matter.

Example prompt

A high-quality 3D render of a cute fluffy monster eating a giant donut; the fur simulation is incredibly detailed, the donut glaze is sticky and reflective, bright daylight lighting, shallow depth of field.

Start with a practical brief

A short poster headline

Swap in your subject and keep the structure.

Design a clean exhibition poster. Exact headline: "FORM AND LIGHT". One ceramic sculpture, off-white background, large dark headline at the top, no other words.

Simple packaging

Swap in your subject and keep the structure.

Create a front-facing coffee bag concept. Exact brand text: "NORTH". Cream paper, one blue mountain illustration, one centered readable word. No extra text.

Before you keep the result
  • Read every word and punctuation mark.
  • Check margins and alignment.
  • Shorten the text if the layout is crowded.

When to switch models

If typography is not central to your brief, compare Pro for general visual work.

Resolution tiers
Standard
Reference images
Up to 4

Find the model for your next image

AI image models

Compare your image options

Compare the settings available for your next image.

SettingsFlux.2 ProFlux.2 MaxFlux.2 FlexFlux.2 KleinSeedream 5.0 LiteSeedream 5.0 ProNano Banana 2GPT Image 2.5
Resolution tiersStandardStandardStandardStandard2K / 3K / 4K1K / 2K0.5K / 1K / 2K / 4KSource size when editing
Quality levelsStandardStandardStandardStandardStandardStandardStandardLow / Medium / High
Reference imagesUp to 4Up to 4Up to 4Up to 4Up to 4Up to 4Up to 4Up to 4
Images per generation11111111
Base credits17 credits38 credits38 credits10 credits14 credits54 credits32 credits16 credits
Base credits by tier2K: 14 / 3K: 14 / 4K: 141K: 27 / 2K: 540.5K: 24 / 1K: 32 / 2K: 48 / 4K: 63Low: 8 / Medium: 16 / High: 32

The Generate button shows the cost for your selected settings and references.

Questions about Flux.2 Flex

Inputs, editing, and credits—before you start.

Can I change inference steps here?

No. The editor uses preset generation settings. You can choose framing and work with prompts and references.

What can I make with Flux.2 Flex?

Generate an image from a written brief, edit one uploaded image, or combine references. Start with a subject, setting, lighting and the details that matter to you.

How many references can I use?

FluxContext supports up to 4 references for this model. Upload your own JPG, PNG or WebP images, up to 4 MB each.

Which settings can I change?

Available choices: Standard. Choose your framing and add references when you want to guide the result.

How many credits does an image use?

Pricing uses fixed tiers for settings and reference images. The Generate button shows the cost of your image.

Can I keep editing a result?

Yes. Choose Edit on a delivered image to use it as your next input. Describe the change and what should stay the same. Each new generation has its own credit charge.

Do I need an account?

This model requires sign-in and enough credits. For a free first experiment, use Flux.2 on the main generator.

Will it preserve every detail?

Not always. Review faces, product details, spelling and composition before using an image. An editing instruction is guidance, not a guarantee that every other pixel remains unchanged.

Create your next image with Flux.2 Flex

Describe your idea and start creating.

Start creating