🎨 Seedream 5.0 Pro — ByteDance's flagship image model: up to 2K, best-in-class typography & multi-reference editing!

Try It!
Logo
Image ModelsExplore All AI Models

Unlimited Ultra-Fast FLUX.2 Klein Text to Image / Image Edit Generator (4B/9B)

Black Forest Labs' fastest AI image model family with 4B and 9B variants. Generate stunning images in under a second with unified text-to-image and image editing capabilities. Runs on consumer GPUs with as little as 13GB VRAM.

Flux Model Series

FLUX.2 Klein Text to Image Showcase

Experience the blazing fast sub-second image generation with FLUX.2 Klein 4B and 9B variants

Try Now
Teotihuacán Pyramids (4B)

Teotihuacán Pyramids (4B)

Ancient pyramids with futuristic Saturn in the sky - cinematic depth and atmospheric perspective

“ancient monumental Teotihuacán pyramids illuminated by red lights, in the sky a giant Saturn suspended in the sky, sculptural and non-astronomical, with rings of liquid metal, futuristic yet archaic, evoking a sacred portal rather than a planet”

FLUX.2 Klein Image Edit Showcase

Transform your images with FLUX.2 Klein's intelligent editing capabilities in under a second

Try Now
Comic Style (4B) - 1
Comic Style (4B)

Comic Style (4B)

01/10

What's FLUX.2 Klein

Black Forest Labs' fastest compact model family for sub-second image generation and editing

  1. 4B/9BParameters
  2. <1sGeneration
  3. 13GBMin VRAM
  4. 4-7Credits

FLUX.2 Klein is a revolutionary compact model family that unifies text-to-image generation and image editing in a single architecture, delivering state-of-the-art quality in under a second.

Key Features

  1. 01

    Sub-Second Generation

    Generate stunning images in under a second with step-distilled 4-step inference, enabling real-time creative workflows and interactive applications.

  2. 02

    Unified Architecture

    Single model architecture supporting text-to-image generation, single-reference editing, and multi-reference composition without switching between different models.

  3. 03

    Consumer GPU Ready

    Run locally on consumer GPUs with as little as 13GB VRAM for the 4B model, making professional-grade AI accessible on RTX 3090 and RTX 4070.

  4. 04

    Multi-Reference Support

    Combine multiple input images with text prompts for context-aware editing and generation, perfect for style transfer and character consistency.

  5. 05

    Dual Size Options

    Choose between 4B for maximum speed or 9B for enhanced quality, both delivering exceptional results that match or exceed models 5x their size.

  6. 06

    Open Weight Models

    4B model available under Apache 2.0 license for maximum flexibility, enabling custom deployments, fine-tuning, and commercial applications.

  7. 07

    NVIDIA Optimized

    FP8 and NVFP4 quantized variants provide up to 2.7x faster inference with 55% less VRAM, built in collaboration with NVIDIA for RTX GPUs.

  8. 08

    Rectified Flow Technology

    Built on cutting-edge rectified flow transformer architecture with Qwen3 text embedder, finding the straightest path between noise and image for optimal quality.

How to Use FLUX.2 Klein Text to Image

Generate images in under a second with the fastest FLUX model

  1. 01

    Write Your Prompt

    Enter a detailed description of the image you want to create. Include subject, style, lighting, mood, and composition for best results. FLUX.2 Klein supports both 4B and 9B variants.

  2. 02

    Choose Model Variant

    Select between FLUX.2 Klein 4B (faster, 13GB VRAM) or 9B (higher quality, 29GB VRAM) based on your hardware and quality requirements.

  3. 03

    Generate Image

    Click generate and FLUX.2 Klein produces your image in under a second. The 4-step distilled inference delivers state-of-the-art quality with unprecedented speed.

FLUX.2 Klein YouTube Videos

Watch demonstrations and tutorials showcasing Black Forest Labs FLUX.2 Klein's ultra-fast AI image generation and editing capabilities

  • FLUX.2 Klein First Test – Fast LOCAL Image Generation & Editing! - Bijan Bowen
  • Flux 2 Klein just released and it runs on a potato! - Sebastian Kamph
  • Flux.2 Klein: Fast Image Editing & Generation in ComfyUI - Code Crafters Corner
  • FLUX.2 Klein Explained 🔥 Fast AI Images & Editing on Low VRAM - AI Ninja
  • NEW FLUX.2 Klein: AI Images in 0.5s 🤯 - Julian Goldie SEO

FLUX.2 Klein YouTube Videos

Watch demonstrations and tutorials showcasing Black Forest Labs FLUX.2 Klein's ultra-fast AI image generation and editing capabilities

FLUX.2 Klein Popular Reviews on X

See what people are saying about FLUX.2 Klein on X (Twitter)

AI Pricing Plans

Flexible AI Pricing

Pay-as-you-go credits or subscription plans. No hidden fees, cancel anytime.

One Time supports crypto payment (BTC, USDT, ETH, 350+) · Monthly billing

Free

Try before you buy

Free
32points
  • Up to 3 videos
  • Up to 32 images
  • Multi-Model Support
  • Text to Video
  • Image to Video
  • Video to Video
Show more (+11)
  • Consistent Character
  • AI Animation Generator
  • Templates & Effects
  • AI Video Enhancers
  • Interactive Community
  • Faster Generation Speed
  • No-watermark Outputs
  • More Camera Movement
  • Private Video Visibility
  • Copy Protection
  • Priority Support

Pro

Popular

Elevate your AI experience

29.99USD / 1 Month
800
800 credits/mo
  • Up to 80 videos
  • Up to 800 images
  • 3 tasks(Parallel Tasks)
  • Multi-Model Support
  • Text to Video
  • Image to Video
Show more (+12)
  • Video to Video
  • Consistent Character
  • AI Animation Generator
  • Templates & Effects
  • AI Video Enhancers
  • Interactive Community
  • Faster Generation Speed
  • No-watermark Outputs
  • More Camera Movement
  • Private Video Visibility
  • Copy Protection
  • Priority Support

Lite

Start your AI journey

19.99USD / 1 Month
300 credits/mo
  • Up to 30 videos
  • Up to 300 images
  • 3 tasks(Parallel Tasks)
  • Multi-Model Support
  • Text to Video
  • Image to Video
Show more (+12)
  • Video to Video
  • Consistent Character
  • AI Animation Generator
  • Templates & Effects
  • AI Video Enhancers
  • Interactive Community
  • Faster Generation Speed
  • No-watermark Outputs
  • More Camera Movement
  • Private Video Visibility
  • Copy Protection
  • Priority Support
  • Card
  • Crypto accepted (BTC, USDT, ETH, 350+)
View Detailed Pricing

FLUX.2 Klein Frequently Asked Questions

Still have questions?

[email protected]DiscordJoin our DiscordSubmit a Ticket
FLUX.2 Klein is Black Forest Labs' fastest AI image model family, designed for sub-second generation and editing on consumer GPUs. Unlike larger models like FLUX.2 Pro or FLUX.2 Dev, Klein uses a compact architecture with 4B and 9B parameter variants that run on hardware like RTX 3090 or RTX 4070 with as little as 13GB VRAM. It uniquely unifies text-to-image generation and image editing in a single model, eliminating the need to switch between different tools.
The 4B model is optimized for maximum speed, generating images in as little as 0.3 seconds on high-end hardware, requiring only ~13GB VRAM, and is released under Apache 2.0 license for commercial use. The 9B model delivers higher quality results matching models 5x its size, targets ~29GB VRAM (RTX 4090 and above), and takes about 0.5-2 seconds per image. Both use the same 4-step distilled inference and share the unified generation/editing architecture.
FLUX.2 Klein supports three core capabilities in a single model: text-to-image generation from prompts, single-reference image editing where you modify an existing image with instructions, and multi-reference generation where multiple input images and a prompt jointly define the output. This makes it perfect for rapid prototyping, style transfer, character consistency, and interactive creative applications.
FLUX.2 Klein 4B text-to-image costs 4 credits per image, while the 9B variant costs 5 credits. For image editing, the 4B model costs 5 credits and 9B costs 7 credits per edit. This makes Klein one of the most affordable options for high-quality AI image generation, especially considering its sub-second speed and professional-grade output quality.
Yes, FLUX.2 Klein is designed for local deployment on consumer GPUs. The 4B model runs on GPUs with 13GB VRAM (RTX 3090, RTX 4070), while the 9B model requires ~29GB VRAM (RTX 4090 and above). Additionally, FP8 and NVFP4 quantized versions reduce VRAM requirements by up to 55% while maintaining quality. The 4B model is Apache 2.0 licensed, allowing commercial use.
The Distilled variants (default on Dreamega) use 4-step inference optimized for speed, perfect for production and interactive use. Base variants are undistilled, preserving the complete training signal for maximum flexibility. Use Base when you need inference step control, plan to fine-tune with LoRA adapters, or want higher output diversity for research and custom pipelines.

Privacy preferences

We use essential storage to run the service. Optional analytics, advertising and convenience features stay off until you choose to enable them. Privacy policy