FLUX.2 klein 9B KV (Image to Image) GGUF

Note: All model files and custom nodes below must be downloaded via the client. If you do not have the client, Download >>>

FLUX.2-klein-9b-kv is a specialized variant optimized with KV-cache support for lightning-fast creative workflows. The "KV" feature acts as a computational shortcut: it stores essential data from reference images during the first step, allowing the model to skip redundant calculations in subsequent edits. This makes multi-image referencing and iterative editing up to 2.5x faster than standard models. For everyday users, this translates to a "real-time" experience where high-quality image adjustments happen almost instantly as you refine your prompts.

FLUX.2-klein-9B-KV GGUF >9 files

Name	Size	Target Dir
Flux-2-Klein-9B-KV-Q2_K.gguf	3.57 GB	unet
Flux-2-Klein-9B-KV-Q3_K_S.gguf	4.46 GB	unet
Flux-2-Klein-9B-KV-Q3_K_M.gguf	4.56 GB	unet
Flux-2-Klein-9B-KV-Q4_K_S.gguf	5.62 GB	unet
Flux-2-Klein-9B-KV-Q4_K_M.gguf	5.72 GB	unet
Flux-2-Klein-9B-KV-Q5_K_S.gguf	6.71 GB	unet
Flux-2-Klein-9B-KV-Q5_K_M.gguf	6.81 GB	unet
Flux-2-Klein-9B-KV-Q6_K.gguf	7.87 GB	unet
Flux-2-Klein-9B-KV-Q8_0.gguf	9.98 GB	unet

Qwen3 8B GGUF >5 files

Name	Size	Target Dir
Qwen3-8B-Q4_K_M.gguf	5.03 GB	text_encoders
Qwen3-8B-Q5_0.gguf	5.72 GB	text_encoders
Qwen3-8B-Q5_K_M.gguf	5.85 GB	text_encoders
Qwen3-8B-Q6_K.gguf	6.73 GB	text_encoders
Qwen3-8B-Q8_0.gguf	8.71 GB	text_encoders

FLUX.2 VAE >1 files

Name	Size	Target Dir
flux2-vae.safetensors	336.21 MB	vae