Skip to content
Other

ToonOut

background removal for anime art

1 min readupdated

ToonOut is a specialized fine-tune of BiRefNet built specifically for removing backgrounds from anime-style images. Standard segmentation models handle realistic photos well, yet they often lose fine details on stylized artwork—such as thin hair strands, line art, and semi-transparent areas. Trained on a dedicated dataset of 1,228 images, this model generates high-precision masks for anime characters. It is a solid fit for artists, editors, and developers working with anime content and graphics.

Key features

  • High accuracy: Accuracy on the test set jumped from 95.3% to 99.5%.
  • Tailored for anime: Focused on preserving fine hair details, line art, and transparency.
  • Solid foundation: Built on top of the proven BiRefNet architecture.
  • Open license: Released under the MIT license.

What you can do

  • Clean up artwork: Remove backgrounds from anime illustrations and characters while keeping complex outlines intact.
  • Plug into pipelines: Swap out standard segmentation algorithms for this model in your anime graphics workflows.

Example

You can run and test the model using the official demo notebook in the project's GitHub repository.

Read moreDetails
Architecture: BiRefNet (fine-tuned)
Pixel Accuracy: 99.5%
Training: On the ToonOut dataset (1,228 images)
License: MIT
Discuss Telegram 0 29

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 0.8 GB · ready files1.4 GB
  • FP8 / INT8recommendedweights 0.4 GB1.0 GB
  • 4-bitweights 0.2 GB0.8 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.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · BF16 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

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.

    # 2524a/toonout
    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 2524a/toonout --local-dir ~/models/toonout
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to toonout
  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.

  • 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 (2)
    • 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.
  • 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 BiRefNet (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 9All on Hugging Face

About the model

2524a/toonout
Author
2524a
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

290

Ask AI about this model

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

Comments