EdgeSVG
raphaelmansuy.github.io
- Free plan
- Windows
- macOS
- Linux
- Web
What EdgeSVG is
EdgeSVG is a free, open-source tool that converts raster images into SVG vector graphics. Auto mode classifies an image as a logo, icon, illustration, or photo and selects a tracing strategy. Other modes include convert, logo, premium, smart, and optimal; smart mode searches iteratively within SSIM and file-size constraints. Quality scoring can report SSIM, edge similarity, foreground IoU, topology, file size, and path counts. The project provides Rust, Python, Node.js, command-line, and WebAssembly interfaces under a shared SDK contract. Its WebAssembly engine runs in the browser without a server or uploads, according to the maker. Batch processing handles directories and produces JSON or Markdown reports with per-image scores and summaries. CI/CD examples cover GitHub Actions, GitLab CI, Docker, and pre-commit hooks. EdgeSVG is published under Apache 2.0, with Python wheels for macOS arm64 and x64, Linux arm64 and x64, and Windows x64. It is not a hosted API or GUI design tool.
Who it suits
EdgeSVG suits designers, developers, and build teams automating vector conversion, batch asset processing, browser conversion, or CI/CD workflows. It is aimed at technical pipelines rather than hosted design software.
Good to know, and what to watch
- Runs vectorization in the browser without uploads
- Offers automatic image-type classification
- Provides multiple tracing modes
- Includes batch reports with quality scores
- Supports Rust, Python, Node.js, CLI, and WebAssembly
- Not a GUI design tool
- Not a hosted API
- Quality modes and scoring require technical workflows
- No server-side processing in browser mode
The verdict
EdgeSVG provides local and programmable raster-to-SVG conversion with several tracing strategies and measurable output scores. Its SDKs, batch tools, and CI/CD examples make it suitable for automated asset pipelines.
What the maker publishes
- Automatic mode
- Auto mode classifies images as logos, icons, illustrations, or photos and selects a tracing strategy. source
- Automatic mode
- Auto mode classifies images as logos, icons, illustrations, or photos and selects a tracing strategy. source
- Automation
- The CI/CD guide documents GitHub Actions and GitLab CI workflows, JSON output, and configurable quality thresholds. source
- Batch processing
- Its benchmark command can process directories and produce JSON and Markdown reports with per-image scores and summaries. source
- Batch processing
- The benchmark command processes directories of raster images and can produce JSON and Markdown reports. source
- Browser privacy
- The maker says the WebAssembly engine runs in the browser without a server or uploads, so image data stays on the device. source
- Browser privacy
- The WebAssembly engine runs in the browser without a server, and the site says image data never leaves the device. source
- Conversion modes
- Documented modes include auto, logo, premium, and smart, with smart mode searching iteratively under SSIM and file-size constraints. source
- Dependencies
- The maker describes EdgeSVG as a pure Rust engine with no GPU, ONNX, or Python ML runtime requirements. source
- Dependencies
- The engine is written in Rust and the site says it has no ML dependencies, GPU requirement, or model downloads. source
- Integrations
- The CI/CD guide provides examples for GitHub Actions, GitLab CI, Docker, and pre-commit hooks. source
- License
- The project is published under the Apache 2.0 license. source
- License
- The project site identifies EdgeSVG as open source under the Apache 2.0 license. source
- Maker
- The site copyright names Raphael Mansuy. source
- Not a hosted service
- The project README says EdgeSVG is not a GUI design tool or hosted API. source
- Platform support
- The Python quick start lists pre-built wheels for macOS arm64 and x64, Linux x64 and arm64, and Windows x64. source
- Purpose
- EdgeSVG converts raster images into SVGs through an open-source vectorization engine. source
- Purpose
- EdgeSVG converts raster images into SVG files. source
- Quality scoring
- It scores vectorized output with metrics including SSIM, edge similarity, foreground IoU, topology, file size, and path counts. source
- Quality scoring
- It reports more than 15 quality metrics for generated SVGs, including SSIM, edge F1, topology, file size, and path counts. source
- SDKs
- EdgeSVG provides Python, Node.js, Rust, CLI, and WebAssembly interfaces with a shared response contract. source
- SDKs and CLI
- The project provides Rust, CLI, Python, Node.js, and WebAssembly surfaces with a shared SDK contract. source
- Supported Python platforms
- The Python quick start lists pre-built wheels for macOS arm64 and x64, Linux x64 and arm64, and Windows x64. source
- Target users
- The maker describes use cases including design pipelines, asset automation, web applications, and build pipelines. source
Checked against the maker's pages on 3 Oct 2026.