FLUX.1 Kontext [dev] VRAM Requirements
FLUX.1 Kontext [dev] needs about 12.3 GB of VRAM with FP8 weights at 1024×1024, loading one model at a time. In FP16 it needs 23.3 GB (35.4 GB with everything loaded), and GGUF Q4 brings it to 7.5 GB. The smallest consumer GPU that runs it is the RTX 3050 8GB with Q4 weights, or the RTX 4060 Ti 16GB in FP8.
12.3 GB peak VRAM
FLUX.1 Kontext [dev] at 1024×1024 with FP8 weights, loading one model at a time (ComfyUI default on smaller cards). Keeping everything loaded needs 19.7 GB.
- Diffusion model 11.1 GB
- Text encoders 4.5 GB
- VAE 0.2 GB
- Peak activations 3.1 GB
- Overhead 0.8 GB
GPU fit
- RTX 4060 8GB61% on GPU
- RTX 3060 12GB92% on GPU
- RTX 4060 Ti 16GBFitswith swapping
- RTX 4090 24GBFitsall loaded
- RTX 5090 32GBFitsall loaded
- RTX 6000 Ada 48GBFitsall loaded
- RTX PRO 6000 Blackwell 96GBFitsall loaded
- Apple M4 Max 128GBFitsall loaded
The VAE decode is the peak here. Turn on tiled VAE decoding to cut it to about 0.8 GB.
FLUX.1 Kontext [dev] VRAM by precision
At 1024×1024. Text encoders in FP8 except for the FP16 row.
| Weights | Diffusion model | Peak (swapping) | All loaded | Smallest tier |
|---|---|---|---|---|
| FP16 / BF16 | 22.2 GB | 23.3 GB | 35.4 GB | 32 GB |
| FP8 | 11.1 GB | 12.3 GB | 19.7 GB | 16 GB |
| GGUF Q8_0 | 11.8 GB | 13.0 GB | 20.4 GB | 16 GB |
| GGUF Q5_K_S | 7.6 GB | 8.8 GB | 16.3 GB | 10 GB |
| GGUF Q4_K_S / NF4 | 6.3 GB | 7.5 GB | 15.0 GB | 8 GB |
VRAM by resolution
FP8 weights, with model swapping.
| Resolution | Latent tokens | Activations | Peak VRAM |
|---|---|---|---|
| 1024×1024 | 4,096 | 0.4 GB | 12.3 GB |
| 1536×1536 | 9,216 | 0.8 GB | 12.7 GB |
About FLUX.1 Kontext [dev]
An instruction-based image editor built on the FLUX.1 [dev] architecture: you give it an image and a text instruction. Weights are the same size as FLUX.1 dev, but the input image is fed in as extra tokens, so activation memory is roughly double at the same resolution. GGUF and FP8 versions run in ComfyUI on 12 GB cards.
| Component | Size |
|---|---|
| Diffusion model | 11.9B parameters (Rectified-flow transformer (image editing)) |
| Text encoders | CLIP ViT-L (0.123B), T5-XXL encoder (4.76B) |
| VAE | 0.084B parameters |
| Classifier-free guidance | Distilled (1 pass per step) |
| Released | 2025-06 |
| License | FLUX.1 [dev] Non-Commercial License |
FLUX.1 Kontext [dev] VRAM: common questions
Does FLUX.1 Kontext need more VRAM than FLUX.1 dev?
Slightly. The weights are identical, but the reference image doubles the number of tokens the transformer processes, which raises activation memory, most noticeably at high resolutions.
How much VRAM does FLUX.1 Kontext [dev] need?
At its default 1024×1024: about 23.3 GB in FP16 with model swapping (35.4 GB with everything loaded), 12.3 GB in FP8 and 7.5 GB with GGUF Q4 weights.