● Covered by 2 sources · 2 reportsMedium impact2 neutral
Black Forest Labs Releases FLUX.3 Image Klein Model in Diffusers for Open Image Generation
Key points
- FLUX.3 Image Klein is an open image generation model.
- It supports text-to-image and multi-reference image-to-image.
- Distilled checkpoint generates 1024x1024 images in ~2 seconds on H100.
- Includes LoRA training scripts and modular Diffusers pipeline.
- FLUX.3 Image Klein is a 5.6B parameter model.
- It supports native 4K generation and pixel-perfect editing.
- The model is optimized for consumer GPUs.
- It includes FP8, INT8, and NVFP4 builds.
- FLUX.3 Image Klein is released under the FLUX Kommunity License.
New Open Image Generation Model
Black Forest Labs has introduced FLUX.3 Image Klein, an open image generation model, now available through the Hugging Face Diffusers library. This release includes two open checkpoints within the FLUX.3 Image Klein collection, providing tools for various image generation tasks.
Key Capabilities and Performance
The FLUX.3 Image Klein model supports text-to-image generation, multi-reference image-to-image transformations, and bounding-box layout control. A distilled checkpoint of the model can generate a 1024x1024 pixel image in approximately two seconds when run on an H100 GPU. The model also offers NVFP4 and FP8 weights for compatibility with smaller GPUs, loaded via `from_single_file`.
Developer Features and Integration
The release includes LoRA training scripts for both text-to-image and image editing functionalities. FLUX.3 Image Klein ships as a Modular Diffusers pipeline, allowing developers to customize its components by adding or removing blocks. Installation requires updating `diffusers` and `transformers` libraries, with the Qwen3-VL text encoder used for text processing.
Prompting and Customization
FLUX.3 responds best to detailed scene descriptions, encompassing subject, setting, lighting, composition, and style. The distilled checkpoint generates 1024x1024 images in 16 steps by default, with options to adjust height, width, or the number of inference steps. Users can also pass reference images to guide image generation, allowing for modifications to existing images based on new prompts.
Updates
- FLUX.3 Image Klein is a 5.6B parameter model.
- It supports native 4K generation and pixel-perfect editing.
- The model is optimized for consumer GPUs.
- It includes FP8, INT8, and NVFP4 builds.
- FLUX.3 Image Klein is released under the FLUX Kommunity License.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Primary sources
GitHub huggingface/blog GitHub ostris/ai-toolkit GitHub huggingface/diffusers GitHub black-forest-labs/flux-3-image-klein GitHub NVIDIA/TensorRT-LLM GitHub NVIDIA-NeMo/AutomodelHow outlets covered it
FLUX 3 Image Klein, a 5.6B parameter open-weights model, has been released for text-to-image generation, image editing, and multi-reference generation. This model offers native 4K generation and pixel-perfect editing, and is optimized for consumer GPUs, enabling local execution and fine-tuning on smaller hardware.
Black Forest Labs has released FLUX.3 Image Klein, an open image generation model, within the Hugging Face Diffusers library. This model offers text-to-image, multi-reference image-to-image, and bounding-box layout control, with a distilled checkpoint generating 1024x1024 images in approximately two seconds on an H100 GPU.