Skip to main content
Black Forest FLUX.2 spans the full spectrum of image generation—from sub-second inference with [klein] to highest quality with [max]. Generate photorealistic images with precise control over colors, poses, and composition, or edit existing images by referencing up to 10 sources simultaneously. Choose [klein] for real-time, high-volume generation, [pro] for production at scale, [flex] for fine-grained control, or [max] for maximum quality and grounding search.
Want to try first? Test FLUX.2 [max], [pro], and [flex] in our playground. [klein] is available via our API and on Hugging Face.

​
What Can You Do?

Combine elements from multiple images while maintaining identity across complex scenes. Create ad variants with consistent faces, product mockups in any context, or fashion editorials where models stay consistent.

Fashion editorial: 8 consistent characters from reference images

Character + pose guidance combined

Input references used for the scene above

​
Which Model to Choose?

[klein][max][pro][flex][dev]
Best forReal-time, high-volumeHighest quality, final assetsProduction at scaleQuality with controlLocal development
Multi-referenceUp to 4Up to 8 (API), 10 (playground)Up to 8 (API), 10 (playground)Up to 8 (API), 10 (playground)Recommended max 6
ControlsStandardStandardStandardAdjustable steps & guidanceFull customization
Grounding searchNoYesNoNoNo
Pricingfrom $0.014 / imagefrom $0.07 / MPfrom $0.03 / MP$0.06 / MPFree (non-commercial)
FLUX.2 [klein] delivers sub-second inference with open weights. 4B runs on consumer GPUs (~13GB VRAM). Apache 2.0 for 4B, FLUX NCL for 9B. See model details below.
FLUX.2 [max] includes grounding search: when prompted, it performs web searches to access real-time information to visualize trending products, current events, or the latest styles without manually sourcing reference material.

​
Compare FLUX.2 Models

​
At a Glance

[klein]

Sub-second inference. Our fastest models with open weights. Runs on consumer GPUs (~13GB VRAM). From $0.014/image via API, or run locally with Apache 2.0 (4B) / FLUX NCL (9B).

[max]

Maximum performance. Highest editing consistency across tasks. Vast world knowledge. Strongest prompt following and faithful style representation.

[pro]

Top performance at affordable price. The high quality, production-grade image editing and generation model.

[flex]

Specialized for typography. Best for text rendering and preserving small details.

​
Use Cases

Use CaseFLUX.2 [klein]FLUX.2 [max]FLUX.2 [pro]FLUX.2 [flex]
Product MarketingBulk catalog generation, A/B testing variantsHighest quality hero shots for marketplacesCreate ads at scale for social campaignsText overlay while preserving details
Movie MakingRapid storyboarding, concept explorationTop quality cinematic pre-visualizationRapid ideation and static movie bannersIntros, credits, static advertising
Creative PlatformsCost-efficient generation for all tiersPremium model for highest-tier subsHigh quality backbone at scaleSpecialized text placement
E-commerceHigh-volume product variations, thumbnailsPremium product photographyProduction-grade catalog imagesPrice tags, labels, descriptions
Editorial & FashionRapid mood boards, style explorationFinal hero imagesCampaign imagery at scaleText-heavy layouts

​
FLUX.2 [klein] Models

FLUX.2 [klein] is our fastest model family, delivering state-of-the-art quality with sub-second inference. Unifying generation and editing in a single compact architecture, [klein] is built for applications requiring real-time image generation—and runs on consumer hardware with as little as 13GB VRAM.
FLUX.2 [klein] photorealistic examples

Photorealistic quality

FLUX.2 [klein] diverse output examples

Diverse styles and subjects

Open weights available: [klein] 4B is fully open under Apache 2.0. [klein] 9B is available under the FLUX Non-Commercial License. Download from Hugging Face.

​
API Models

[klein] 4B[klein] 9B
Best forHigh volume, local deploymentBalanced quality and speed
Architecture4B flow model9B flow model + 8B Qwen3 text embedder
Inference steps4 (step-distilled)4 (step-distilled)
VRAM~13GB~24GB
SpeedSub-secondSub-second
API Pricing0.014+0.014 + 0.001/MP0.015+0.015 + 0.002/MP
LicenseApache 2.0FLUX Non-Commercial License

​
Open Weights (Community)

The Base variants are undistilled foundation models with full training signal—ideal for fine-tuning, LoRA training, research, and custom pipelines. Higher output diversity than distilled models.
[klein] Base 4B[klein] Base 9B
Best forFine-tuning, research, custom pipelinesMaximum quality, research
Output diversityHighHighest
Step-distilledNo (full capacity)No (full capacity)
LicenseApache 2.0FLUX Non-Commercial License
AvailabilityHugging FaceHugging Face
Base models are available as open weights for local development and research. They are not offered on the public API.
FLUX.2 [klein] does not include prompt upsampling. Write detailed, descriptive prompts for best results. See our prompting guide for techniques.

​
Technical Specifications

Resolution

  • Output: Up to 4MP
  • Input: 64x64 minimum
  • Any aspect ratio

Multi-Reference

  • Up to 10 input images ([klein]: 4)
  • Character consistency
  • Style transfer

Advanced Controls

  • Pose guidance
  • Hex color matching
  • Structured prompting
  • Grounding search ([max] only)

​
Getting Started

Try in Playground

Test FLUX.2 [max], [pro], and [flex] in your browser. No setup required.

Download [klein] Weights

Get [klein] weights from Hugging Face for local inference.

Text-to-Image API

Generate images from text prompts.

Image Editing API

Edit images with multi-reference support.

[klein] Prompting Guide

Master narrative prompting for best [klein] results.

Local Development

Download [dev] weights for local inference.