Model reference · open weights
BFL-FLUX.2-klein is an open-weight image model from Runware. BFL-FLUX.2-klein-9B (BF16) weighs 34.7 GB; the smallest configuration that runs it is RTX 5090 32 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
What it is
| Released by | Runware |
|---|---|
| Released | 2026-01-19 |
| Parameters | 3.9B |
| VRAM | 34.7 GB for the weights |
What it runs on
| Card | One image | Memory |
|---|---|---|
| RTX 3060 12 GB … RTX 4090 24 GB | does not fit | |
| RTX 5090 32 GB | fits (encoders offloaded) | 31.0 GB |
| L40S 48 GB | tight | 44.0 GB |
| A100 80 GB | fits | 78.2 GB |
| H100 80 GB | fits | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | fits | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | fits | 107 GB |
| H200 141 GB | fits | 138 GB |
| B200 180 GB | fits | 176 GB |
Builds
| Build | Params | Precision | Weights | Smallest setup |
|---|---|---|---|---|
| BFL-FLUX.2-klein-4B ↗ | 3.9B | BF16 | 7.9 GB | RTX 4060 Ti 16 GB |
| BFL-FLUX.2-klein-9B (above) ↗ | 9.1B | BF16 | 34.7 GB | RTX 5090 32 GB |
Running it yourself
Rent a machine by the hour. Runs as it is with diffusers, on the machine, in Python.
# on your rented machine: pip install diffusers transformers accelerate
import torch
from diffusers import DiffusionPipeline
from diffusers.utils import load_image
pipe = DiffusionPipeline.from_pretrained("Runware/BFL-FLUX.2-klein-4B", torch_dtype=torch.bfloat16).to("cuda")
start = load_image("/workspace/in.png")
image = pipe(prompt="the same scene at golden hour", image=start).images[0]
image.save("/workspace/out.png")