FLUX.2 [klein] 9B on 8 GB of VRAM
Not at full quality. The 16-bit and 8-bit files are too big for 8 GB; the largest that fits is Q3_K_M (4.8 GB). A distilled 9B FLUX.2 for fast generation and editing in a few steps. Much lighter than FLUX.2 dev.
Q3_K_M (4.8 GB) fits with only 0.9 GB to spare (about 7.1 GB of 8 GB). Quality: noticeable loss of detail. For better quality, Q4_K_M (5.9 GB) also runs, with about 0.2 GB spilling into system RAM - a little slower, still practical.
Also try the FP8 file (9.4 GB) on NVIDIA GPUs. It does not fit entirely, but a current ComfyUI streams the overflow from RAM and native 8-bit files often beat a small GGUF that does fit. Dynamic VRAM →
01Other amounts of VRAM
02Every file, on 8 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| 16-bit SAFETENSORS · black-forest-labs | 18.2 GB | 20.5 GB | 12.5 GB too big | the original weights | Hugging Face → |
| Q8_0 GGUF · unsloth | 10.0 GB | 12.3 GB | 4.3 GB too big | practically identical to the original | Hugging Face → |
| FP8 SAFETENSORS · black-forest-labs | 9.4 GB | 11.7 GB | 3.7 GB too big | practically identical to the original | Hugging Face → |
| Q6_K GGUF · unsloth | 7.9 GB | 10.2 GB | 2.2 GB too big | very close to the original | Hugging Face → |
| Q5_K_M GGUF · unsloth | 7.0 GB | 9.3 GB | spills 1.3 GB | close; small differences in fine detail | Hugging Face → |
| Q4_K_M GGUF · unsloth | 5.9 GB | 8.2 GB | spills 0.2 GB | good; some loss in fine detail and text | Hugging Face → |
| Q3_K_M ← GGUF · unsloth | 4.8 GB | 7.1 GB | fits · 0.9 GB spare | noticeable loss of detail | Hugging Face → |
| Q2_K GGUF · unsloth | 4.0 GB | 6.3 GB | fits · 1.7 GB spare | heavy loss; a last resort | Hugging Face → |
“Needed” adds 1.5 GB of working memory for FLUX.2 klein 9B and a 0.8 GB reserve. Sizes 2026-10-01.
03What to download
| Part | File | Folder | Size | |
|---|---|---|---|---|
| Model | flux-2-klein-9b-Q3_K_M.gguf Q3_K_M · unsloth/FLUX.2-klein-9B-GGUF | models/unet | 4.8 GB | Download → |
| Text encoder | qwen_3_8b_fp8mixed.safetensors Qwen3 8B FP8 · Comfy-Org/flux2-klein-9B | models/text_encoders | 8.7 GB | Download → |
| VAE | flux2-vae.safetensors FLUX.2 VAE · Comfy-Org/flux2-dev | models/vae | 0.3 GB | Download → |
| Total download · keep about the same free on disk | 13.8 GB | |||
Alternatives: Qwen3 8B BF16 (16.4 GB). Sizes from Hugging Face, 2026-10-01.
Free ComfyUI workflow, tested on my RTX 5060 Ti: GGUF Q8_0 workflow (.json). Drag the file into ComfyUI. All workflows →
04The text encoder
At 8.7 GB even the smallest Qwen3 8B file is too big for 8 GB, so it runs from system RAM. That slows down reading the prompt, not the picture. Sizes: 16-bit 16.4 GB, FP8 8.7 GB. Why the encoder does not count against VRAM →