Can the RTX 5090 32 GB run FLUX.2 [dev]?
Download Q6_K (27.4 GB). With the model's working memory it needs about 30.7 GB, leaving 1.3 GB spare on 32 GB. Quality: very close to the original.
Worth trying on this GPU: with an up-to-date ComfyUI, Dynamic VRAM (on by default for NVIDIA since March 2026) streams whatever does not fit from system RAM, and ComfyUI's own start-up message recommends native FP8/INT8 files over GGUF, saying they “will be faster even if they are larger than your memory”. So before settling for a heavily compressed GGUF, try the native FP8 file (35.5 GB). The file recommended above is the best one that fits entirely — the safe choice on older ComfyUI versions, AMD and Intel. Comfy blog: Dynamic VRAM →
01What to download
| Part | File | Folder | Size | |
|---|---|---|---|---|
| Model | flux2-dev-Q6_K.gguf Q6_K · city96/FLUX.2-dev-gguf | models/unet | 27.4 GB | Download → |
| Text encoder | mistral_3_small_flux2_fp8.safetensors Mistral 3 Small FP8 · Comfy-Org/flux2-dev | models/text_encoders | 18.0 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 | 45.8 GB | |||
Alternatives: Mistral 3 Small BF16 (35.6 GB). The 16-bit text encoder is a little more faithful; the smaller one is chosen here because it loads faster and needs less RAM. Sizes read from Hugging Face (2026-09-25). “Download” links start the file directly; the file name links are the same files the official ComfyUI workflows use.
System RAM: 64 GB or more. ComfyUI keeps the model file, the text encoder and the VAE in system RAM and moves them to the GPU as needed. With this set of files that is about 45.8 GB, plus roughly 6 GB for Windows, ComfyUI and a browser: 51.8 GB in total. With less RAM it still runs, but Windows starts swapping to disk and loading gets very slow. calculated
02Every FLUX.2 dev file on 32 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| 16-bit SAFETENSORS · black-forest-labs | 64.4 GB | 67.7 GB | 35.7 GB too big | the original weights | Hugging Face → |
| FP8 SAFETENSORS · Comfy-Org | 35.5 GB | 38.8 GB | 6.8 GB too big | practically identical to the original | Hugging Face → |
| Q8_0 GGUF · city96 | 35.0 GB | 38.3 GB | 6.3 GB too big | practically identical to the original | Hugging Face → |
| Q6_K ← GGUF · city96 | 27.4 GB | 30.7 GB | fits · 1.3 GB spare | very close to the original | Hugging Face → |
| Q5_K_M GGUF · city96 | 24.1 GB | 27.4 GB | fits · 4.6 GB spare | close; small differences in fine detail | Hugging Face → |
| Q4_K_M GGUF · city96 | 20.1 GB | 23.4 GB | fits · 8.6 GB spare | good; some loss in fine detail and text | Hugging Face → |
| Q3_K_M GGUF · city96 | 16.0 GB | 19.3 GB | fits · 12.7 GB spare | noticeable loss of detail | Hugging Face → |
| Q2_K GGUF · city96 | 12.9 GB | 16.2 GB | fits · 15.8 GB spare | heavy loss; a last resort | Hugging Face → |
Sizes from the Hugging Face file listing, read 2026-09-25. “Needed” = file + 2.5 GB working memory for this model + 0.8 GB kept free for the system.
03The text encoder
Mistral Small 3 (24B): 35.6 GB as 16-bit, 18.0 GB as FP8. ComfyUI encodes the prompt first and can push the encoder out of VRAM before sampling, so it does not have to fit together with the model. On 32 GB the FP8 encoder fits on its own, so prompt encoding stays fast.
04About this GPU
The RTX 5090 32 GB is a Blackwell GPU with FP8 and FP4 hardware: ComfyUI computes Comfy-Org's FP8 files natively here, and NVFP4 files (where a model offers them) are faster still.
RTX 5090 32 GB: 32 GB GDDR7 · 512-bit · 1792 GB/s · Blackwell. Everything that runs on the RTX 5090 32 GB →
05What more VRAM would change
With 48 GB you could run FLUX.2 dev at 8-bit or better (FP8, 35.5 GB) with no offloading.
06Measured and reported results
Nobody has sent measured numbers for this pair yet. If you run FLUX.2 dev on a RTX 5090 32 GB, send your time per image and peak VRAM and it will appear here, credited.
07Questions
How much VRAM does FLUX.2 dev need?
Around 16.2 GB with the smallest file (Q2_K) and 38.8 GB with an 8-bit file (FP8), counting working memory and a small system reserve. The full 16-bit file needs about 67.7 GB.
Which FLUX.2 dev file should I download for the RTX 5090 32 GB?
Q6_K (27.4 GB) from city96/FLUX.2-dev-gguf. Load it with Unet Loader (GGUF) from the ComfyUI-GGUF node pack.
Is FP8 faster than GGUF on the RTX 5090 32 GB?
It can be. This GPU has FP8 hardware, and ComfyUI computes FP8 natively for files made for it (Comfy-Org's fp8_scaled files), or for any FP8 file with the --fast fp8_matrix_mult option. GGUF files are unpacked on the fly, which costs some speed.
How much do I need to download for FLUX.2 dev on the RTX 5090 32 GB?
About 45.8 GB for the model file, text encoder and VAE listed on this page, and the same again free on disk. With a 32 GB card, 64 GB of system RAM or more is recommended.