Skip to content
Images

FLUX.2 [dev]

a powerful new image generation and editing model

2 min readupdated

FLUX.2 [dev] is a 32-billion-parameter rectified flow transformer model built for text-guided image generation, editing, and composition. Out of the box, it transfers characters, objects, and styles without any fine-tuning. Trained using guidance distillation, it delivers higher overall efficiency. The project targets researchers, artists, and developers looking to build innovative visual workflows.

Key features

  • State-of-the-art text-to-image generation alongside single- and multi-reference editing.
  • Zero fine-tuning required for precise character, object, and style transfer.
  • Higher efficiency thanks to guidance distillation.
  • Open weights for advancing research and running commercial or personal creative projects under a non-commercial license.

What you can do

  • Generate detailed images using complex text descriptions and prompts.
  • Use multiple reference images simultaneously for precise control over characters and objects in the frame.
  • Integrate the model into popular interfaces like ComfyUI or run it locally using the Diffusers library.
  • Conduct generative AI research and build advanced creative pipelines.

Example

import torch
from diffusers import Flux2Pipeline

repo_id = "diffusers/FLUX.2-dev-bnb-4bit"
device = "cuda:0"

pipe = Flux2Pipeline.from_pretrained(
    repo_id, text_encoder=None, torch_dtype=torch.bfloat16
).to(device)

prompt = "Realistic macro photograph of a hermit crab using a soda can as its shell..."
image = pipe(generator=torch.Generator(device=device).manual_seed(42)).images[0]
image.save("flux2_output.png")
Read moreDetails
Architecture: rectified flow transformer (32 billion parameters).
Tools: full support in Diffusers and ComfyUI.
License: FLUX [dev] Non-Commercial License.
Safety: built-in input and output filters to prevent unwanted content generation.
Discuss Telegram 0 140

Will it run?

Runs in 4-bit

Full precision does not fit, a quantised version does: quality barely changes down to 8 bits, more noticeably at 4 bits.

  • BF16weights 60 GB · ready files68 GB
  • FP8 / INT8recommendedweights 30 GB35 GB
  • 4-bitweights 17 GB20 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # siyarajput/FLUX.2-dev-Backup
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download siyarajput/FLUX.2-dev-Backup --local-dir ~/models/FLUX.2-dev-Backup
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to FLUX.2-dev-Backup
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • SageAttention 3v2.2.0from October 28, 2025

    even faster on RTX 50 / Blackwell

    FP4 attention for Blackwell. Faster than SageAttention 2, but the authors suggest v2 when accuracy matters.

    How to enable: Build from the sageattention3_blackwell folder of the repo; in ComfyUI via KJNodes.

    Earlier generations (2)
    • SageAttention 2++ — Quantised (INT8/FP8) attention with no visible quality loss. The main speed-up for image and video generation.
    • SageAttention 1 — First version: INT8 for Q·K. Still works, but 2++ is faster and more accurate.
  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (3)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
    • xFormers — The older way to speed up attention, works almost everywhere. Sage or Flash are faster where available.
  • cache-dit (DBCache)v1.5.6from November 20, 2026new release

    1.5–3× at nearly the same quality

    Reuses block outputs between diffusion steps. Supports most DiT models.

    How to enable: pip install cache-dit; one line cache_dit.enable_cache(pipe) for diffusers.

    Earlier generations (2)
    • MagCache — Skips steps by magnitude of change. More accurate than TeaCache, no calibration.
    • TeaCache — The first popular diffusion step cache. Works, newer methods are more accurate.
  • Nunchaku (SVDQuant 4-bit)v1.2.1from January 25, 2026

    ≈3× faster and 3.5× less memory

    4-bit versions of popular image models with almost no quality loss (FP4 on RTX 50).

    How to enable: Ready weights for supported models; in ComfyUI use ComfyUI-nunchaku.

  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of FLUX.2-dev-Backup (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

siyarajput/FLUX.2-dev-Backup
Author
siyarajput
Category
Images
Released
Posted

On NeyroLearn

Hugging Face

1400

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments

Other

Iris-Z

Iris-Z