FLUX.2 [klein]

Generate and edit in less than a second with state-of-the-art quality. Available now via our API, or fine-tune and run it locally.

Blazing fast.

Sub-second inference*, making it 30%+ faster than any competing model.

* Inference time varies by model and hardware. See model comparison for more details.

Production-quality visuals as fast as you can prompt.

We built a free demo so you don't have to take our word for it.
No signup, no credit card.

Access [klein] through our platform in the way that works best for you.

Integrate FLUX.2 [klein] directly into your applications with our production-ready API.

Try FLUX.2 [klein] instantly in our interactive playground. No setup required.

FLUX.2 [klein] runs on your hardware.
Choose from four variants optimized for different use cases.

Model	Description	License	Inference Time (GB200)In seconds	Inference Time (RTX5090)In seconds	VRAM
FLUX.2 [klein] 9B	Our distilled model. Outstanding quality at sub-second speed. Great for real-time generation while retaining quality. Marketing launch will focus on this model.	FLUX Non-Commercial License	~0.5	~2	19.6 GB
FLUX.2 [klein] 9B Base	Our undistilled foundation model. Maximum flexibility and control. Great for fine-tuning.	FLUX Non-Commercial License	~6	~35	21.7 GB
FLUX.2 [klein] 4B	The fastest variant in the Klein family. Built for interactive applications, real-time previews, and latency-critical production use cases.	Apache 2.0	~0.3	~1.2	8.4 GB
FLUX.2 [klein] 4B Base	A smaller foundation model with exceptional quality-to-size ratio. Ideal for local deployment, fine-tuning on limited hardware, and efficient generation and editing workflows.	Apache 2.0	~3	~17	9.2 GB