Is this model really open source?

Yes. Black Forest Labs released FLUX.2 [klein] 4B under the Apache 2.0 license, and here it runs as a hosted API call with nothing to install.

Does it refuse NSFW image prompts?

No. Nothing is reviewed or refused, and no request is turned into a different one. The limit is the model: in our text-to-image tests it often added garments on its own, while edits delivered the Moderate level and kept an unfiltered source as asked. The Very strong level came back toned down in both modes; the Mild and Maximum-level prompts rendered.

Why did it add clothes I did not ask for?

A 4B model has limited range, and it leans towards covered figures. Describe the outfit you want, or choose a larger open model if you need more.

How is it different from the larger FLUX.2 models?

It is much smaller and quicker, and it generates and edits in one model. Text, colors and faces are less reliable than on larger models. Choose it for drafts, mockups and fast edits.

Can I edit with more than one picture?

Yes, up to three. Name them in the prompt as image 1, image 2 and image 3 and say what each contributes. Faces can drift slightly when three sources are combined.

How do aspect ratio and size work in edits?

Auto keeps the shape of the first source image at about one megapixel. Choosing a ratio lets you pick the 1k or 2k size tier.

Can I get the same image twice?

Set a seed to keep runs comparable while you change one detail of the prompt. The output may still vary a little.

Am I charged for a failed request?

No. Failed requests are not charged.

FLUX.2 Klein 4B API: Fast Uncensored NSFW Images

black-forest-labs/flux-2-klein-4b/edit

FLUX.2 [klein] 4B image edit: change or combine up to three source images with a written instruction, in the same small open model, with no content filter.

From $0.013/image

Run model View API

Input

Image editing

Prompt338/5000

Reordering updates prompt references. Removing media also removes its references; review your prompt.

Images1/3

Drop files or media links, paste, or browse · Up to 3 files · Max 10 MB · JPG · PNG · WEBP · GIF

image 1

Aspect ratio

Resolution

Seed

Safety checkerOff

SpicyAPI does not run a safety checker on this model. Your prompt and media go to the model as written, and the model itself decides what comes back.

Checking your account before enabling submission and pricing.

Output

Ready for a new run

Example result

Animate

Edit

Split into layers

{
  "model": "black-forest-labs/flux-2-klein-4b/edit",
  "input": {
    "image_urls": [
      "https://cdn.spicyapi.ai/models/examples/inputs/8504e34119d44313.jpeg"
    ],
    "aspect_ratio": "auto",
    "resolution": "1k",
    "seed": 260160,
    "prompt": "Keep the couple from image 1 exactly as they are: same faces, clothes, pose and framing. Replace the sunny street with a rain-soaked street at night: wet asphalt reflecting magenta and blue neon, light rain falling, and a glowing sign above them that reads \"NIGHT MARKET\". Warm light from the shop windows on their faces. Realistic photo."
  }
}

Introduction

FLUX.2 [klein] 4B Image edit changes or combines up to three source images with a written instruction, on Black Forest Labs' smallest FLUX.2 model, which is open source under Apache 2.0. Keep the shape of your first image with auto, or choose a ratio and a 1k or 2k size tier. Each edit is one flat price per image, and no content filter is applied.

Choose a mode by starting material

What you have	Mode
a written description	FLUX.2 Klein 4B Text to image
one to three pictures to change or combine	FLUX.2 Klein 4B Image edit

Key features & highlights

Up to three references

Combine two portraits and a location, or restyle a single photo. The price does not change with the number of sources.

Edits that keep people

Ask for a new background, light or weather and keep faces, clothes and pose. With one source this works cleanly; with three, faces can drift slightly.

Open, with no filter

Nothing is reviewed or refused. In our tests edits kept an unfiltered source as it was; only the Very strong level came back toned down.

Request parameters

Request parameters for FLUX.2 Klein 4B · Image Edit
Parameter	Required	Description
`prompt`	Yes	Say what should change and what must stay the same, up to 5,000 characters. With several sources, refer to them as image 1, image 2 and image 3. at least 1 characters · at most 5000 characters
`image_urls`	Yes	One to three source images, up to 10 MB each, in the order your prompt refers to them. Required. at least 1 items · at most 3 items
`aspect_ratio`	No	Auto (default) keeps the shape of the first source at about one megapixel. Or pick one of nine ratios from 21:9 to 9:21. `auto` · `1:1` · `16:9` · `9:16` · `4:3` · `3:4` · `3:2` · `2:3` · `21:9` · `9:21` · Default: `auto`
`resolution`	No	1k (default) or 2k, which is 2048 pixels on the long side. It only applies when you choose a ratio; with auto the size is about one megapixel. `1k` · `2k` · Default: `1k`
`seed`	No	A number from 0 to 2147483647. Reuse it to compare edits while you change one detail of the prompt. `≥ 0` · `≤ 2147483647`

Pricing

$0.013 / image

Billed per image at one flat price, whatever the number of source images or the size tier. Failed requests are not charged.

Use cases

Background and mood swaps

Move a couple from a sunny street to a rainy neon night in one pass.

Bringing people to a place

Place two portraits in a location photo for storyboards, covers and character sheets.

Quick restyles

Change light, season or palette while the composition stays put.

Explore key use-case tutorials

Prompt guide & pro tips

Protect first, change second

Open with what must remain the same, then describe only the change. Short, concrete instructions work best.

Give each source a role

With several images, say what each one contributes: this person, that person, this place.

Check the details

With two or three sources, look at hands, hair and accessories; a small extra item can appear. Run another seed if something is off.

API quick start

# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
  -H "Authorization: Bearer $SPICY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
  -d '{
  "model": "black-forest-labs/flux-2-klein-4b/edit",
  "input": {
    "aspect_ratio": "auto",
    "image_urls": [
      "https://example.com/reference.png"
    ],
    "prompt": "A ceramic teapot on a linen cloth, soft window light, muted palette",
    "resolution": "1k"
  }
}'

# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
  -H "Authorization: Bearer $SPICY_API_KEY"

See the API reference for authentication, the complete schema, polling, and error handling.

Missing the model you need?

Tell us what to add and where you would use it. Requests go to the people who decide what gets onboarded next.

OPEN-SOURCE IMAGE API

FLUX.2 Klein 4B API at a glance

FLUX.2 [klein] 4B: the smallest, fastest FLUX.2 model, Apache 2.0 open source, with no content filter.

Parameters: 4B; Black Forest Labs' smallest FLUX.2 model, released as Apache 2.0 open source.
Typical time per image: 15–35 s; What our example runs took; the larger FLUX.2 models usually need longer, and busy periods add queue time.
Reference images: 3; Edit or combine up to three source images in one request.
Aspect ratios: 9; From 21:9 to 9:21, at about one megapixel or up to 2048 pixels on the long side.
Content filter: None; No review of the prompt or the result, so nothing is refused for its subject.

Results

What you can make with FLUX.2 Klein 4B

Color and signs
Neon palettes from hex codes
Two women at a midnight roller rink, hot pink light across the floor and a neon sign on the wall. The pink followed its code, while the teal landed on their clothes instead of the lights.
Text to image, 3:4 at 2k, hex colors and a neon sign.
Portrait
Gritty light, real texture
A boxer between rounds under a single tungsten bulb: sweat, grey at the temples and deep shadows in a 2:3 frame at 2K.
Text to image, 2:3 at 2k.
Layout
A poster with a headline
A perfume ad with a serif headline and a small tagline, both spelled correctly. The burgundy backdrop came out darker than the code asked for.
Text to image, 9:16 at 1k, quoted headline and tagline.
Edit
New street, same couple
One daytime photo became a rainy neon night with a new sign overhead, while faces, clothes and the arm around the shoulder stayed as they were.
Image edit with one source, aspect ratio auto.
Three references
Two portraits and a place
Two portraits and a beach cabin merged into one sunset porch scene. Faces stay close to the sources, with a little drift in the first man's hair.
Image edit with three sources, 16:9 at 2k.

Content test

Is FLUX.2 Klein 4B uncensored? What we tested

We ran each level on our live setup and checked every clip frame by frame. A pass means the content was actually on screen, not just that the request went through.

Content levels tested on FLUX.2 Klein 4B, by mode
Level	Image editing	Text to image
MildThe lightest prompts in the test	RendersTested Oct 8, 2026	RendersTested Oct 8, 2026
ModerateWhere filtered models start to refuse	RendersTested Oct 8, 2026	Not tested yet
StrongThe middle of the scale	Not tested yet	Not tested yet
Very strongWhere most models start to hold back	Not tested yet	Not tested yet
MaximumThe top of the scale	RendersTested Oct 8, 2026	RendersTested Oct 8, 2026

Renders: on screen as written. Softened: generated, but milder than what was written. Refused: rejected. Not tested yet: we have not run that level on that mode.

How each mode behaves

All modes: No prompt or image is refused and no filter touches the result. In text to image, though, this is the most conservative model in our tests: it often adds garments on its own to Moderate and Strong prompts. Edits held back less: the Moderate level came through as asked and an unfiltered source stayed as it was. The Very strong level came back toned down in both modes, while the Mild and Maximum-level prompts did render.
All modes: Text inside images and hex colors are not reliable at this size. In one test, three of five tries misspelled the same neon sign. Faces drift slightly when three sources are merged, and hands are often wrong with two people in the frame, so run a few seeds and pick the best.
All modes: Up to three source images per edit and about 2048 pixels on the long side. There is no negative prompt, no step count and no guidance setting here. Output is JPEG.

Use cases

What to build with the FLUX.2 Klein 4B API

Fast drafts and iteration
At seconds per image it suits trying many prompts, seeds and framings before you commit to a slower, larger model.
Posters and signs with neon light
Quoted headlines, small taglines and hex-coded palettes work well enough for ad mockups, event posters and moody night scenes.
Quick background and mood swaps
Hand it one photo and ask for a new street, light or weather while faces, clothes and pose stay put.
Putting people and places together
Feed in portraits and a location and place everyone in one frame, for storyboards and character sheets.
Open-model work without filters
The model has no moderation layer, which fits platforms and projects that need an open model to run as an ordinary API call.

Prompts

FLUX.2 Klein 4B prompt examples you can copy

Paste one into the playground as it is, then change one thing at a time.

Text to image
Roller rink in neon
Two women in their thirties at a retro roller rink after midnight, skating hand in hand, one pulling the other close by the waist, both laughing. One wears a cropped satin bomber jacket over a lace bralette and high-waisted shorts; the other wears a fitted off-shoulder ribbed top, a short pleated skirt and knee socks. Hot pink (#FF3EA5) and teal (#2EE6D6) neon light washes across the polished wooden floor, a mirror ball hanging from the ceiling scatters specks of light, and a neon sign on the back wall reads "SKATE NIGHT". Flash photography, motion blur in the background.
Why it works. Outfits are described piece by piece, so the model has no room to improvise. Colors come with hex codes and the sign text is quoted. Expect to try a few seeds for clean lettering.
Try this prompt
Text to image
Boxer between rounds
An adult man in his forties, a shirtless boxer in a basement gym between rounds, wearing red leather boxing gloves and black satin boxing shorts, resting both forearms on the top rope of the ring and looking straight into the camera. Sweat on his shoulders and chest, short dark hair with grey at the temples, short stubble. A single hanging tungsten bulb, dust in the air, a worn heavy bag in the background, deep shadows. Gritty, high-contrast color photograph, 50mm lens.
Why it works. One light source, a few physical details and a lens give the frame a gritty, photographic feel. Gloves cover the hands, which helps at this model size.
Try this prompt
Text to image
Perfume poster
Vertical perfume advertisement poster. A woman in her thirties with auburn hair in a loose pinned updo, wearing a backless deep-red velvet evening gown, seen from behind as she glances back over her shoulder, holding a faceted crystal perfume bottle near her collarbone. Dark burgundy (#4A0E1E) backdrop, soft gold rim light along her shoulders and spine. An elegant serif headline at the top reads "NOCTURNE" and a small line at the bottom reads "Eau de Parfum".
Why it works. Layout is spelled out: a headline at the top, a small line at the bottom and the subject in between. Short words in quotes are the safest text to ask for.
Try this prompt
Image editing
Swap the street, keep the couple
Keep the couple from image 1 exactly as they are: same faces, clothes, pose and framing. Replace the sunny street with a rain-soaked street at night: wet asphalt reflecting magenta and blue neon, light rain falling, and a glowing sign above them that reads "NIGHT MARKET". Warm light from the shop windows on their faces. Realistic photo.
Why it works. It starts with what must stay unchanged, then describes only the new setting and the sign. The faces and clothes carry over because the prompt protects them.
This mode starts from your own image: add it in the playground before you run.
Try this prompt
Image editing
Merge three references
Place the man from image 1 and the man from image 2 together on the front porch of the wooden beach cabin from image 3 at sunset. Both now wear unbuttoned linen shirts, white on the man from image 1 and sand-colored on the man from image 2, with rolled-up light trousers, barefoot. The man from image 1 sits on the porch railing; the man from image 2 stands close beside him with an arm around his waist, both smiling. Keep both faces exactly as in their photos. Golden backlight from the low sun, the ocean behind them, realistic photo.
Why it works. Each source gets a clear role by number: two people and one place. Outfits and poses are written out, and the closing sentence asks to keep both faces.
This mode starts from your own image: add it in the playground before you run.
Try this prompt

Pricing

FLUX.2 Klein 4B API pricing

One image starts at $0.0093 at 1K and $0.0099 at 2K.

The playground shows the exact price for your settings before you run. Failed or timed-out tasks are usually not charged. Some models charge for content refusals as stated on their model page; check those billing terms before submitting.

See every price tier

Get started

How to use the FLUX.2 Klein 4B API in three steps

Step 1
Quote the text you want
Put any words for signs, posters or labels in quotation marks and say where they sit. Short words come out well; a 4B model can still misspell, so try a few seeds.
Step 2
Name colors as hex codes
Write a code such as #FF3EA5 next to the thing it colors. Expect the right hue rather than an exact match, and note that it may land on clothes instead of light.
Step 3
Say what stays in an edit
In edits, state what must stay the same (faces, clothes, pose, framing) before describing the change. With three sources, refer to them as image 1, image 2 and image 3.

Open the playground Get an API key

Parameter reference and code samples are in the API and Readme tabs at the top of this page.

Why SpicyAPI

Why run FLUX.2 Klein 4B on SpicyAPI

No filter from us
We add no content filter of our own. What the model renders is what you get back.
Tested, not guessed
Content labels come from real generations we ran and checked, with the test date on the page.
Pay per result, in dollars
No subscription and no credits. Failed or timed-out tasks are usually not charged. Some models charge for content refusals as stated on their model page; check those billing terms before submitting.
One key for every model
One API key and balance cover image, video, audio and chat models, so you can explore more models from the same account.

FAQ

FLUX.2 Klein 4B API FAQ

About FLUX.2 Klein 4B

FLUX.2 [klein] 4B is Black Forest Labs' smallest FLUX.2 model, open source under Apache 2.0, with text to image and up to three reference images in one model, nine aspect ratios and no content filter. It is quick and good for drafts, posters and background swaps, but it often dresses people more than you asked and cannot render the Very strong level.