flux.2 takes an image you already have — usually one z-image-turbo just generated — and edits it: a new style, a color shift, a composition tweak, without starting the prompt from zero.
Your generated image will appear here
flux.2 is an image-to-image model: instead of starting from a blank prompt, it takes an existing picture — almost always one z-image-turbo just generated — and produces a modified version of it. Where z-image-turbo answers “what should this image look like from scratch,” flux.2 answers “how should this specific image change.” That distinction is the whole reason BytePlus AI runs two separate models rather than one: generation and editing are genuinely different tasks, and a model tuned for one usually isn't the best choice for the other.
On this site, flux.2 sits between the two halves of the hero tool. You generate a base image with z-image-turbo, and if the composition is right but the style, color, or a detail needs work, flux.2 is the model built to make that change without throwing away what already worked.
Edit, don't restart
Rewriting a text prompt from scratch to fix one detail almost always changes the whole image — the composition, the pose, the background all shift with it. flux.2 is built to make a targeted change instead: a different color palette, a new art style, a lighting adjustment, while keeping the underlying image recognizable as the same one you started with.
Where it fits in the workflow
The typical loop is: write a prompt, let z-image-turbo generate, then decide whether to keep it, regenerate, or hand it to flux.2 for a restyle. Only after that is the result a candidate for the animate-to-video preview step described on our Seedance-alternative page.
Image to Image
Takes an existing picture as input, not a blank prompt.
0 Sign-Ups
Edit and restyle without creating an account.
100% Free to Try
No credits, no paywalled first edit.
The most common request is a style transfer: turning a photographic result into a watercolor painting, a flat illustration, or a clay-render look, without redoing the composition. Close behind that is color and lighting correction — warming up a cold-looking scene, deepening shadows, or shifting a daytime shot toward golden hour, all changes that are hard to describe accurately in a fresh text prompt but straightforward as an edit instruction against an existing image.
flux.2 is also useful for detail refinement: sharpening a product shot's texture, cleaning up an awkward edge, or nudging a small element without regenerating the whole scene. And because it starts from a real image rather than noise, it's well suited to producing a small family of variations on one idea — useful when you've found a composition that works and want a few different treatments of it rather than five unrelated new images.
In practice most people never think of these as two separate products — they just generate, look at the result, and decide whether it needs a restyle before moving on. If you want to try the pairing directly, start at the free AI image generator or the hero tool on the homepage, generate with z-image-turbo, and use the result as your flux.2 starting point.
Independence note: flux.2 is a hosted third-party model that BytePlus AI integrates, not something we trained ourselves, and byteplusai.site is not affiliated with the official BytePlus company. See BytePlus AI Explained for the full picture.
Yes. Editing an image with flux.2 is free, with no account and no credit card — the same anonymous, no-sign-up model that covers every generation tool on this site.
No. flux.2 works from the image itself, not the prompt that created it. You describe the change you want — a style, a color shift, a detail — and it applies that to the picture you already have.
Style transfers (photo to watercolor or illustration), color and lighting adjustments, and small detail fixes tend to work best. Very large structural changes, like completely re-posing a subject, are usually better served by regenerating with z-image-turbo instead.
No. flux.2 is an image-editing model only. The animate-to-video step on this site is a separate, honestly-labeled motion preview — see the Seedance-alternative page for exactly what that does and doesn't do.
They're both hosted models that BytePlus AI integrates, run back to back in the same free tool, but they are separate models built for different jobs — generation versus editing.
Other models we run and explain.
Transactional pages built around the same models.
Start with z-image-turbo, then refine the result — free, anonymous, and ready right now.
byteplusai.site is an independent AI creation playground and is not affiliated with, endorsed by, or sponsored by ByteDance or BytePlus Pte. Ltd. "BytePlus" is referenced solely to describe the AI model category we help you explore.