FLUX 3 Image

Create composed images from a prompt or references, then refine details in the Oeneo AI workbench.

Reference images (Optional)

0/10
0/5000

Model

Aspect ratio

Resolution

Output format

FLUX 3 Image examples

How does FLUX 3 Image change one detail, merge references or render 4K?

Explore targeted photo edits, multi-reference composition and native 4K image generation. The before-and-after edits and the 4K crops are BFL official examples; the multi-reference image illustrates the same workflow.

The same portrait with a cat graphic added to the white T-shirt.
Original portrait of a smiling woman wearing a plain white T-shirt.
Original
Edited

Targeted image editing

Change one element, keep the rest of the photo

Describe the single change you want, for example adding a print to a shirt, and state that everything else stays. After each edit, check the face, pose and background, since repeated edits can drift.

The living room with a blue sofa replacing the foreground armchairs.
Original living room with patterned armchairs, a rug and wood-paneled walls.
Original
Edited

Interior design variations

Try furniture options in a real room photo

Start from a photo of the actual room and describe only the furniture swap, so walls, windows and flooring stay as reference points. Compare the before and after to see how the new piece changes the room before you buy or pitch it.

A boxy retro hatchback painted in a yellow pattern of blue squiggles, red triangles and black dots takes a hairpin bend on a sunny alpine pass road; corner cards show the three references: the plain white car in a parking lot, the painted pattern swatch and the empty mountain road.

Multi-reference composition

Combine several references into one scene

Give each reference a role, such as the subject, a material or the setting, and FLUX 3 Image blends them with shared lighting and scale. You can upload up to 10 references; check which details from each source carried through.

Official native 4K soba-shop image above two detail crops showing the Japanese shop lettering and a diner behind glass.

Native 4K image generation

Generate at native 4K when small text and textures matter

Native 4K means the model renders the full resolution itself instead of upscaling afterwards, which matters when signs, materials or small figures must hold up in print or tight crops. The crops under this 5,456 × 3,072 official example come from the same file.

Which design briefs suit FLUX 3 Image?

Beyond edits and references, FLUX 3 Image suits design work where element placement and lettering decide whether the image is usable.

Aerial view of two models whose vast pleated red gowns spread like discs on a dirt road between banana trees.

Composition-aware generation

Posters, covers and editorial pages where the position of every element matters. BFL designed FLUX 3 Image to reason about how elements share a frame, so placement and negative space in your prompt carry real weight.

A research-paper page titled “Advancements in Steamed Ham Technology”, with an abstract, two text columns, a data table and a line chart.

Multilingual text rendering

Headlines, labels and document-style pages with readable lettering, in more than one language.

How do you describe layout, edits and references to FLUX 3 Image?

Describe the result in plain language. For an edit, focus on the change; for a composition, explain how the elements relate.

Make the brief specific

  1. Place the subject and the text

    Specify left or right, foreground or background, and where the title needs clear space. Put exact short text in quotation marks.

  2. Name the change and the parts to preserve

    For example: change the mug to matte black; keep the hands, table and light. Avoid redescribing the entire image.

  3. Assign a role to every reference

    Identify the subject, clothing, material or setting from each image. Keep upload order consistent with the brief.

Leave room to verify

  1. Proofread lettering and product details

    Readable text and reference guidance are capabilities, not guarantees. Inspect names, dates, faces and packaging before use.

  2. Oversized references

    Very large uploads, including a 4K result you want to edit again, can be refused; the upload limits are in the specs below.

  3. Bounding boxes are a separate workflow

    BFL demonstrates a visual layout canvas. This Oeneo AI page does not offer a bounding-box or mask editor.

How does FLUX 3 Image compare with Ideogram 4.5 and GPT Image 2.5?

Compare the current Oeneo AI input capacity, output choices and credit costs. The prices use each model's default quality, one output and one reference for editing. Changing quality or adding references can change the quote.

SpecificationFLUX 3 ImageIdeogram 4.5GPT Image 2.5 FlareGPT Image 2.5 Sunburst
Output resolution tiers1K, 2K, 4K1K, 2K1K, 2K, 4K1K, 2K, 4K
Choose it whenMulti-reference scenes, 4K finishPosters and typography-heavy editsFast exploration, high volumeHigh-stakes single edits
Total input imagesUp to 10Up to 51 source + 4 guide referencesUp to 16Up to 16
Images per request1Up to 4Up to 4Up to 4
Output formatsPNG, JPEGProvider defaultPNG, JPEG, WEBPPNG, JPEG, WEBP
Selectable qualityNo quality selectorlow, medium, highlow, medium, high, xhigh, maxlow, medium, high, xhigh, max
1K generation / editing10 / 10 creditsQuality: provider controlled12 / 12 creditsQuality: medium2 / 3 creditsQuality: low2 / 3 creditsQuality: low
Extra reference costIncluded in the quoteIncluded in the quoteFirst 1 included, then +2 credits eachFirst 1 included, then +2 credits each
Explore the modelFLUX 3 ImageIdeogram 4.5GPT Image 2.5 FlareGPT Image 2.5 Sunburst

FLUX 3 Image resolution tiers, upload limits and credits

Generate from text or upload references to edit. Your selected resolution determines the credit quote; the reference-input limit is separate from output resolution.

FLUX 3 Image resolution tiers, upload limits and credits
Image tasks
Text-to-image generation and reference-based image editing
Editing references
1–10 images, each at least 256 px on both sides and no larger than 4 MP.
Output sizes
1K, 2K, 4K
Output formats
PNG, JPEG. PNG does not imply a transparent background.
Credits per image
1K: 10, 2K: 20, 4K: 122. Generation and editing share these rates; reference images have no additional charge.
Images per run
1

Where FLUX 3 Image stops: video, layout boxes, 4K edits and weights

Is FLUX 3 an image model or a video model?+

FLUX 3 is Black Forest Labs' multimodal model family. FLUX 3 Image handles still-image generation and editing. Earlier FLUX 3 releases and guides also cover video; creating a clip requires a separate video workflow.

What's new in FLUX 3 Image compared with FLUX.2?+

FLUX 3 Image is the newer Black Forest Labs model, released on October 1, 2026. Compared with FLUX.2 Pro it adds native 4K output, takes up to 10 reference images here instead of 8, and is built for edits that change one detail and leave the rest. A 1K image costs 10 credits, against 6 for FLUX.2 Pro and 1–2 for FLUX.2 Klein, so the FLUX.2 models still suit photoreal product shots and low-cost drafts.

How many reference images can I use, and does their order matter?+

You can provide up to 10 images in one FLUX 3 Image editing request. Describe which image supplies each subject, product or setting. With Auto aspect ratio, the first reference determines the output shape, so keep upload order consistent with your instructions.

Can I make several edits to the same image?+

Yes. Download a result and upload it for the next edit, state only the new change, and inspect the parts you want to preserve after each round. This is not an automatic editing session.

Can I draw bounding boxes to control the layout?+

BFL's official Playground demonstrates bounding-box composition and editing. This page does not include that canvas or a mask editor. You can describe positions and relationships in a prompt, but that is different from direct visual layout control.

Can I keep editing a 4K image at its original resolution?+

Not through the current fal endpoint if the image exceeds 4 MP. Output resolution and reference-input limits are separate. Resizing a large image also changes its pixels, so original-resolution, pixel-identical editing should not be assumed.

Can FLUX 3 Image generate Chinese or multilingual text?+

BFL describes multilingual text generation, and its official examples include lettering inside images. Chinese text quality has not been verified in Oeneo AI. Keep copy short and proofread every character, name and date.

Is FLUX 3 Image free on Oeneo AI?+

You can try it with free credits: new accounts get 10, and a 1K FLUX 3 Image costs 10 credits. Larger sizes cost more, as listed in the specs above. Free-plan images are public; paid plans can keep them private.

Are FLUX 3 Image open weights available?+

In its October 1, 2026 launch announcement, BFL described open weights as a future release and commercial weights as available by arrangement. Check BFL for the latest licensing status. Hosted image creation and licensing weights for deployment are separate decisions.

Create a FLUX 3 Image from your brief

Use the FLUX 3 image generator for a new composition, or upload references for a focused FLUX 3 image edit in Oeneo AI.