withoutbg/withoutbg-python

Python SDK for local and cloud background removal (pip install withoutbg)

What it solves

It provides a simple way to remove image backgrounds in Python, allowing users to choose between running a free, private local model or using a high-quality Cloud API for better results on complex details like hair and fur.

How it works

The library offers a unified API for two different execution paths:

  • Local Mode: Uses an open-weights ONNX model (approximately 455 MB) downloaded from Hugging Face that runs on your CPU. It combines depth, segmentation, matting, and refinement in a single pass using a DINOv3 backbone.
  • Cloud Mode: Connects to the withoutBG Cloud API for professional-grade quality without requiring local hardware resources.

Who it’s for

  • Developers building scripts, notebooks, or backends that need automated background removal.
  • Privacy-conscious users who prefer offline, unlimited processing on their own machine.
  • Product builders who need high-precision cutouts for commercial applications via API.

Highlights

  • Dual-path API: Switch between local and cloud processing with a single line of code.
  • CPU-friendly: Local inference runs on CPU via ONNX, meaning no GPU is required.
  • Batch processing: Includes a remove_background_batch method to process multiple images efficiently by loading the model once.
  • PIL Integration: Returns images as PIL Image objects in RGBA mode, making it easy to composite images or save them as PNG/WebP.
  • CLI Tool: Includes a command-line interface for quick background removal without writing code.

Related

  • Project
    danielgatis/rembg

    A tool for removing image backgrounds that can be used as a CLI, Python library, or HTTP server, supporting various AI models and hardware acceleration.

  • Project
    onnx/onnx

    ONNX is an open, framework‑agnostic model format and reference implementation that lets you move AI models between tools and runtimes. Install via `pip install onnx`, use the Python API to build or load a graph, then run inference with any ONNX‑compatible engine.

  • Project
    pangxiaobin/image-matting

    A desktop application that uses the RMBG-1.4 model to provide local AI-powered background removal, batch processing, and image editing tools.

  • Project
    alvarobartt/hf-mem

    hf‑mem is a lightweight Python CLI (and library) that estimates the RAM required to run any Hugging Face model that uses Safetensors or GGUF weights. It fetches only metadata via HTTP range requests, supports Transformers, Diffusers, Sentence‑Transformers, and can optionally estimate KV‑cache memory and MoE breakdowns. Installable as a standalone tool, a Hugging Face CLI extension, or an agent skill.

  • Project
    inisis/OnnxSlim

    A tool for simplifying and slimming ONNX models to reduce operator count and improve inference speed while maintaining accuracy.