license: apache-2.0
base_model: black-forest-labs/FLUX.2-klein-4B
base_model_relation: quantized
pipeline_tag: text-to-image
tags:
- text-to-image
- image-to-image
- flux
- flux2
- flux2-klein
- comfyui
- checkpoint
- safetensors
- aio
- fp8
- nvfp4
language:
- en
🚀 FLUX.2 [klein] 4B AIO
Sub-second image generation — all-in-one, 4 – 6 steps, text-to-image and image editing, Apache-2.0
-7C3AED?style=flat-square){kind=icon}
ComfyUI/models/checkpoints/, load it with Load Checkpoint and generate with 4 – 6 steps and CFG 1.0 (this value is critical).
✨ At a glance
| Property | Details |
|---|---|
| 🧩 Type | All-in-one checkpoint: UNet + Qwen3 text encoder + VAE in a single file |
| 🏗️ Base model | black-forest-labs/FLUX.2-klein-4B (repackaged and converted) |
| 📦 Files | FP8 ( |
| 🎛️ Settings | 4 – 6 steps · CFG 1.0 (critical) · euler · simple or normal · 1024×1024 |
| 🧰 Workflow | ComfyUI workflow AIO-T2I-V1 in workflows/ (Civitai page) |
| 📜 License | Apache-2.0, free for personal and commercial use |
✨ What is FLUX.2 [klein] 4B AIO?
FLUX.2 [klein] 4B AIO is an All-in-One repackage of Black Forest Labs' newest compact image generation model. This version includes VAE, Text Encoder (Qwen3) and UNet in a single file – just load and go!
"Klein" means "small" in German – but this model is anything but limited. It delivers exceptional performance in Text-to-Image, Image Editing and Multi-Reference Generation, typically reserved for much larger models.
💎 Flux.2-4B-Distilled-AIO
The only version on Civitai (4 precision files: FP8, FP16, BF16, NVFP4). All-in-one checkpoint: UNet + Qwen3 text encoder + VAE. Civitai: Flux.2-4B-Distilled-AIO · Workflow: 🎨 AIO-T2I-V1 · ComfyUI folder for all files: models/checkpoints/ (Load Checkpoint)
📦 Available files (from the original description)
| Version | File | Precision | UNet | Text Encoder | VAE | Best for |
|---|---|---|---|---|---|---|
| 🟡 FP8-AIO (~7.7 GB) – Recommended for most users | flux-2-klein-4b-fp8-aio.safetensors |
FP8 | FP8 | FP8 | BF16 | Most users, quick tests, everyday use, lowest VRAM |
| 🔵 FP16-AIO (~15 GB) – For older GPUs | flux-2-klein-4b-fp16-aio.safetensors |
FP16 | FP16 | FP16 | BF16 | Older GPUs (GTX 10xx, RTX 20xx), broadest compatibility |
| 🟢 BF16-AIO (~15 GB) – Maximum quality | flux-2-klein-4b-bf16-aio.safetensors |
BF16 | BF16 | BF16 | BF16 | RTX 30xx/40xx/50xx, professional/commercial work |
| 🔴 NVFP4-AIO (~5.7 GB) – Maximum speed ⚡ | flux-2-klein-4b-aio-nvfp4.safetensors |
NVFP4 | NVFP4 | FP4-mix | BF16 | RTX 50xx only, ultra-fast generations, Blackwell GPUs, low VRAM + extreme performance |
⚙️ Settings (from the original description): 4–6 steps · CFG 1.0 (critical) · euler · simple (or normal) · 1024×1024 native · no negative prompt needed.
This version has no separate version description on Civitai.
🖼️ Flux.2-4B-Distilled-AIO gallery
All images below were posted to the Flux.2-4B-Distilled-AIO version on Civitai.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
🔄 Update – ⚡ Flux2-klein-4B-AIO-NVFP4
Fast generation with Blackwell in just a few seconds — even at 4 steps, and scales nicely with more steps 🚀
🕒 Performance
Prompt executed in 2.08 seconds
██████████████ 4 / 4 steps
~2.65 it/s
✅ Extremely fast ✅ Stable ✅ Great quality for a distilled setup
🖼️ Example Generation
Prompt:
Anime, powerful anime illustration with vibrant dark fantasy colors, one adult woman inspired by Jouryuu, tall imposing presence, long hair flowing dramatically, intense anime eyes, wearing an ornate battle-inspired dress referencing official visuals, heavy fabric and strong silhouette, standing confidently in a ruined temple environment, low-angle camera enhancing dominance, dramatic backlighting with red and violet tones, strong shadows, intense cel-shading, bold anime lineart, intimidating yet elegant presence, correct anatomy, no text, no watermark.
(The Civitai description contains no embedded image for this prompt, so none is shown here.)
💡 This setup is optimized for speed, making it ideal for quick iterations, testing ideas, or just having fun generating without long wait times.
Have fun generating — and as always, thanks for all the feedback and support 🙌✨
🎯 Key Features
- ⚡ 4-6 Step Generation – Sub-second inference on modern hardware
- 📦 All-in-One – No separate VAE/Text Encoder download needed
- 🎨 Unified Architecture – T2I, I2I Editing & Multi-Reference in one model
- 📐 1024×1024 native – Optimized for this resolution
- 💾 Low VRAM – Runs on consumer GPUs with ease
- 📜 Apache 2.0 – Fully open for commercial use!
- 🔧 LoRA-compatible – Base version ideal for fine-tuning
⚙️ Recommended Settings
- Steps: 4-6 (step-distilled, more steps ≠ better)
- CFG: 1.0 ⚠️ CRITICAL!
- Sampler: euler
- Scheduler: simple (or "normal")
- Resolution: 1024×1024 (native)
⚠️ CRITICAL: CFG Must Be 1.0!
This is a distilled model optimized for CFG 1.0. Higher CFG values will produce worse results!
✅ CFG 1.0 = Correct ❌ CFG 3.5+ = Wrong, will look bad
Additional Notes
- 4 – 6 steps are optimal! The model was step-distilled for fast inference
- No negative prompts needed – works but not required
- Natural language prompts – Just describe what you want to see
📥 Installation (ComfyUI)
Quick Start
- Download your preferred version (FP8/FP16/BF16)
- Place in
ComfyUI/models/checkpoints/ - Load with the "Load Checkpoint" node
- Generate!
Folder Structure
ComfyUI/
└── models/
└── checkpoints/
└── flux-2-klein-4b-bf16-aio.safetensors (or fp16/fp8)
🎨 Example Prompts
Photorealistic
A professional photograph of a barista making latte art in a cozy
coffee shop, morning light streaming through windows, shallow depth
of field, shot on Sony A7III
Digital Art
A majestic dragon perched on a crystal mountain peak, aurora borealis
in the background, fantasy digital painting, highly detailed scales,
dramatic lighting
Product Photography
Minimalist product photo of a luxury perfume bottle on white marble,
studio lighting, reflection, commercial photography
💻 Capabilities
✅ What FLUX.2 [klein] 4B can do:
- Text-to-Image (T2I) – High-quality image generation from text
- Image-to-Image (I2I) – Single-reference editing
- Multi-Reference – Multiple input images for controlled transformations
- Text Rendering – Improved text rendering in images
- Photorealistic – Professional photo quality
- Artistic Styles – Diverse artistic styles
⚠️ Limitations:
- Optimized for 1024×1024 (other resolutions possible but not optimal)
- 4B model – less detail than larger models for complex scenes
- Distilled version – less output diversity than base models
🔧 Technical Details
- Parameters: 4 Billion
- Architecture: Rectified Flow Transformer
- Text Encoder: Qwen3-based
- Inference Steps: 4-6 (step-distilled)
- Native Resolution: 1024×1024
- Precision: BF16 / FP16 / FP8 (+ NVFP4 since v1.1)
- License: Apache 2.0
🆚 Comparison: 4B vs 9B
| FLUX.2 [klein] 4B | FLUX.2 [klein] 9B | |
|---|---|---|
| Parameters | 4B | 9B |
| VRAM | ~8-13 GB | ~29 GB |
| GPU | RTX 3090/4070+ | RTX 4090+ |
| Quality | Very Good | Excellent |
| License | Apache 2.0 ✅ | Non-Commercial ❌ |
| Commercial Use | Yes! | No |
→ 4B is perfect for: Consumer hardware, commercial projects, fast iterations
❓ FAQ
Q: Do I need separate VAE/Text Encoder files? No! AIO = All-in-One. Everything is included in a single file.
Q: Can I use this for commercial projects? Yes! The 4B version is licensed under Apache 2.0.
Q: Why only 4-6 steps? The model was step-distilled. More steps won't improve quality.
Q: Why must CFG be 1.0? This is a distilled model optimized for CFG 1.0. Higher values will degrade output quality.
Q: FP8 vs BF16 – What's the difference? FP8 is smaller and faster, BF16 has slightly better quality. For most applications FP8 is sufficient.
Q: Does this work with LoRAs? Yes! Especially the Base version (non-distilled) is ideal for LoRA training.
Q: What's the difference from the 9B version? 9B has better quality but is non-commercial only. 4B is Apache 2.0!
🐛 Troubleshooting
Images look "washed out" or oversaturated
- Check the CFG – it must be 1.0 for the distilled model!
- Use 4-6 steps
Poor text rendering
- Be more specific in your prompt
- Use simple, short text
- Place text requirements at the beginning of the prompt
Colors look off
- Try BF16 version instead of FP8
- Ensure your monitor is properly calibrated
🧩 Workflows
🎨 AIO-T2I-V1 workflow
Civitai: Flux.2-klein-Workflows – AIO-T2I-V1 · Download: Flux.2-T2I-AIO.zip · Flux.2-T2I-AIO-workflow-v2.json
What it does: Step 1 Load Checkpoint (AIO) → optional LoRA (LoraLoader, bypass toggle) → Step 2 prompt (CLIPTextEncode, negative via ConditioningZeroOut) → Step 3 size (EmptyFlux2LatentImage) → Step 4 generate (KSampler) → Step 5 optional upscale pass (ImageScaleBy 1.5× lanczos → VAEEncode → second KSampler at denoise 0.2) with before/after Image Comparer; images are saved with metadata. Built-in guide and resolution-preset notes.
⚙️ Settings (from the JSON): base pass 6 steps · CFG 1 · euler · beta · 1024×1536; upscale pass 4 steps · CFG 1 · euler · simple · denoise 0.2.
📦 Models used
| File | Required? | Download | ComfyUI folder |
|---|---|---|---|
Checkpoint (CheckpointLoaderSimple, preset flux-2-klein-4b-bf16-aio.safetensors) |
required (one of the 4 AIO files) | flux-2-klein-4b-bf16-aio.safetensors · FP16 · FP8 · NVFP4 |
models/checkpoints/ |
LoRA (LoraLoader, placeholder your_lora_here.safetensors) |
optional | any FLUX.2 [klein] 4B LoRA | models/loras/ |
🔌 Custom nodes
| Custom node pack (external) | Nodes used in this workflow | Required? |
|---|---|---|
| rgthree-comfy | Fast Groups Bypasser, Image Comparer |
required |
| comfyui_image_metadata_extension | SaveImageWithMetaData |
required |
Install with ComfyUI-Manager (search the name) or git clone into custom_nodes/, then restart ComfyUI. Everything else is ComfyUI core.
🖼️ AIO-T2I-V1 · result image from the workflow page (1)
The Civitai workflow version has no description text (only the files and images).
🙏 Credits
- Original model: Black Forest Labs
- Architecture: Rectified Flow Transformer
- Text Encoder: Qwen3
- AIO repackage: SeeSee
Official Links
📜 License
- This repository is a repackage (single-file AIO + precision conversions) of black-forest-labs/FLUX.2-klein-4B, which is published under Apache-2.0 – including its transformer, Qwen3 text encoder and VAE components as distributed in that repo. The text encoder is Qwen3-based (Qwen/Qwen3-4B, Apache-2.0).
- Note: FLUX.2 [klein] 9B and FLUX.2 [dev] use the non-commercial FLUX license; they are not part of this repository.
📋 Changelog
v1.1 – January 2026 ⚡
- 🆕 Added NVFP4 AIO variant (RTX 50xx / Blackwell)
- ⚡ Ultra-fast inference (optimized for 4 steps)
- 🧠 Extremely low VRAM usage
- 🎯 Designed for maximum speed while keeping good image quality
v1.0 (January 2026)
- Initial Release
- BF16, FP16 and FP8 versions
- All-in-One with VAE + Text Encoder + UNet
License: Apache 2.0 – Free for personal AND commercial use! 🎉
The fastest open-source image generation model for ComfyUI! ⚡ Download and start creating! 🚀
🧭 More from SeeSee — all models
🧩 Checkpoints & AIO
| Model | Hugging Face | Civitai |
|---|---|---|
| AniSee | SeeSee21/AniSee | Model · Workflows |
| Z-Anime | SeeSee21/Z-Anime | Model |
| Z-Image-Turbo-AIO | SeeSee21/Z-Image-Turbo-AIO | Model · Workflow |
| ERNIE-AIO | SeeSee21/ERNIE-AIO | Model · Workflow |
| Chroma-Anime-AIO | SeeSee21/Chroma-Anime-AIO | Model · Workflow |
| FLUX.2-klein-4B-AIO | SeeSee21/FLUX.2-klein-4B-AIO | Model · Workflow |
| QWEN-Anime | SeeSee21/QWEN-Anime | Model · Workflow |
| SeeMoss-Genesis | SeeSee21/SeeMoss-Genesis | Model · Workflow |
| See-Krea2 | SeeSee21/Krea-2-See-Krea2-Genesis | Model · Workflow |
🎀 LoRAs
| Model | Hugging Face | Civitai |
|---|---|---|
| Illustrious Anime Collection | SeeSee21/Illustrious-Anime-Collection | Model |
| Chroma Semi-Anime | SeeSee21/chroma-Semi-Anime | Model |
| Kitsune Furry Girl | SeeSee21/Kitsune-Furry-Girl-Z-Image | Model |
| Ayase Momo (Dandadan) | SeeSee21/Ayase-Momo-Z-Image | Model |
🧩 Checkpoints & AIO · 🎀 LoRAs · 🌐 Civitai · ▶️ YouTube
Made with ❤️ by SeeSee