Skip to content
SPECIFICTION
Get listed

rembg

github.com

Free plan LinuxAPISelf-hosted
Visit github.com
Spec score 8.2 #38 of 94 in AI Background Removers

What it is

rembg is a free tool for removing image backgrounds. It can run as a command-line program, Python library, HTTP server, or Docker container, with API access and batch processing. The command line handles individual images, folders, watch mode, and RGB24 streams; PNG is the listed export format. Models cover general images, portraits, clothing, anime, and object segmentation, with bria-rmbg as the default local model. Edge refinement options include color decontamination, alpha matting, and ViTMatte. Documentation covers FFmpeg RGB24 output, OpenCV, PIL, NumPy, Docker, and the HTTP API. Installation supports CPU, NVIDIA CUDA GPU, and AMD ROCm backends, and requires Python versions >=3.11 and <3.14. Local ONNX models download on first use into the rembg models directory. An optional withoutBG backend uploads images to its servers, with a documented 20 MB maximum. The tool is MIT-licensed, but model weights have separate licenses; commercial RMBG-2.0 use under the BRIA license requires a paid agreement.

Spec sheet

Purpose
rembg is a tool for removing image backgrounds.Read 9 Oct 2026 · github.com
Free plan
YesRead 23 Sep 2026 · github.com
Integrations
The documentation shows using rembg with FFmpeg RGB24 output, OpenCV, PIL, NumPy, Docker, and its HTTP API.Read 9 Oct 2026 · github.com
Support
The README invites users to support the project by buying the maintainer a coffee or beer.Read 9 Oct 2026 · github.com
API access
YesRead 23 Sep 2026 · github.com
Batch processing
YesRead 23 Sep 2026 · github.com
Cloud option
The optional withoutBG backend sends images to withoutBG servers and has a documented maximum upload size of 20 MB.Read 9 Oct 2026 · github.com
Commercial-use limit
The README says the BRIA license for RMBG-2.0 requires a paid agreement for commercial use.Read 9 Oct 2026 · github.com
Compute options
Installation instructions cover CPU, NVIDIA CUDA GPU, and AMD ROCm backends.Read 9 Oct 2026 · github.com
Edge refinement
Edge options include color decontamination, alpha matting, and ViTMatte for refining soft edges.Read 9 Oct 2026 · github.com
Export formats
PNGRead 23 Sep 2026 · github.com
Licensing
rembg is licensed under MIT, while model weights have separate licenses that may affect commercial use.Read 9 Oct 2026 · github.com
Local storage
Local ONNX models download on first use and are stored under the rembg models directory.Read 9 Oct 2026 · github.com
Models
Available models include options for general images, portraits, clothing, anime, and object segmentation; bria-rmbg is the default local model.Read 9 Oct 2026 · github.com
Processing
The command line supports single images, batch folder processing, watch mode, and RGB24 streams.Read 9 Oct 2026 · github.com
Replacement backgrounds
NoRead 23 Sep 2026 · github.com
Runtime requirements
The project lists Python versions >=3.11 and <3.14 as requirements.Read 9 Oct 2026 · github.com
Security detail
Downloaded models have hash verification, which can be disabled with the MODEL_CHECKSUM_DISABLED environment variable.Read 9 Oct 2026 · github.com
Ways to use
It can be used as a command-line tool, Python library, HTTP server, or Docker container.Read 9 Oct 2026 · github.com

Plans and prices

rembg Free MIT-licensed tool · local models downloaded on first use Read 9 Oct 2026 · github.com

Every rembg plan and price →

Strong points and limits

Strong points

  • Free and MIT-licensed tool
  • Offers command line, library, server, and Docker use
  • Supports batch folder processing
  • Provides multiple segmentation models
  • Supports CPU, CUDA, and ROCm

Limits

  • PNG is the listed export format
  • Requires Python >=3.11 and <3.14
  • withoutBG uploads have a 20 MB limit
  • Some model licenses restrict commercial use

Who it suits

rembg suits developers and technical users who need local or service-based background removal through scripts, APIs, containers, or batch workflows. It is useful when PNG output and model or edge-selection options fit the pipeline.

Verdict

rembg provides several integration paths for automated background removal, with batch handling, model choices, and edge refinement. Review the separate model licenses and runtime requirements before commercial or production deployment.

Where Specifiction places it

Alternatives to rembgAll alternatives →