- Home
- Image Vectorization Software
- EdgeSVG
EdgeSVG
What it is
EdgeSVG converts raster images into SVG vector graphics through an open-source vectorization engine. Its automatic mode identifies logos, icons, illustrations, or photos and chooses a tracing strategy. Documented modes include convert, logo, premium, auto, smart, and optimal; smart mode searches iteratively under SSIM and file-size constraints. Output can be scored using SSIM, edge similarity, foreground IoU, topology, file size, and path counts. Interfaces are provided for Rust, Python, Node.js, the command line, and WebAssembly through a shared response contract. The browser WebAssembly engine runs without a server or uploads, according to the maker. Batch commands process directories and create 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 free under Apache 2.0 and supports Linux, macOS, Windows, and web use. The project says it is not a GUI design tool or hosted API.
For and against
For
- Multiple tracing modes
- Quality scoring with several metrics
- Rust, Python, Node.js, CLI, and WebAssembly interfaces
- Browser processing without uploads
- Batch JSON and Markdown reports
Against
- Not a GUI design tool
- Not a hosted API
- Output is SVG only
- No free trial is listed
The verdict
EdgeSVG is a free vectorization engine aimed at automation, development, and design workflows rather than hosted or GUI use. Its mode choices, scoring, browser processing, and batch tooling define its practical scope.
Who it suits
It suits developers and design or build teams automating raster-to-SVG conversion. The listed interfaces and CI/CD workflows also fit web applications, asset pipelines, and batch processing.
On the record
- Automatic mode
- Auto mode classifies images as logos, icons, illustrations, or photos and selects a tracing strategy.raphaelmansuy.github.io · read 3 Oct 2026
- Automation
- The CI/CD guide documents GitHub Actions and GitLab CI workflows, JSON output, and configurable quality thresholds.raphaelmansuy.github.io · read 3 Oct 2026
- Batch processing
- The benchmark command processes directories of raster images and can produce JSON and Markdown reports.raphaelmansuy.github.io · read 3 Oct 2026
- Browser privacy
- The WebAssembly engine runs in the browser without a server, and the site says image data never leaves the device.raphaelmansuy.github.io · read 3 Oct 2026
- Conversion modes
- Documented modes include auto, logo, premium, and smart, with smart mode searching iteratively under SSIM and file-size constraints.raphaelmansuy.github.io · read 2 Oct 2026
- Dependencies
- The engine is written in Rust and the site says it has no ML dependencies, GPU requirement, or model downloads.raphaelmansuy.github.io · read 3 Oct 2026
- Integrations
- The CI/CD guide provides examples for GitHub Actions, GitLab CI, Docker, and pre-commit hooks.raphaelmansuy.github.io · read 2 Oct 2026
- License
- The project site identifies EdgeSVG as open source under the Apache 2.0 license.raphaelmansuy.github.io · read 3 Oct 2026
- Maker
- The site copyright names Raphael Mansuy.raphaelmansuy.github.io · read 3 Oct 2026
- Not a hosted service
- The project README says EdgeSVG is not a GUI design tool or hosted API.github.com · read 2 Oct 2026
- Platform support
- The Python quick start lists pre-built wheels for macOS arm64 and x64, Linux x64 and arm64, and Windows x64.raphaelmansuy.github.io · read 2 Oct 2026
- Purpose
- EdgeSVG converts raster images into SVG files.raphaelmansuy.github.io · read 3 Oct 2026
- Quality scoring
- It reports more than 15 quality metrics for generated SVGs, including SSIM, edge F1, topology, file size, and path counts.raphaelmansuy.github.io · read 3 Oct 2026
- SDKs
- EdgeSVG provides Python, Node.js, Rust, CLI, and WebAssembly interfaces with a shared response contract.raphaelmansuy.github.io · read 3 Oct 2026
- SDKs and CLI
- The project provides Rust, CLI, Python, Node.js, and WebAssembly surfaces with a shared SDK contract.github.com · read 2 Oct 2026
- Supported Python platforms
- The Python quick start lists pre-built wheels for macOS arm64 and x64, Linux x64 and arm64, and Windows x64.raphaelmansuy.github.io · read 3 Oct 2026
- Target users
- The maker describes use cases including design pipelines, asset automation, web applications, and build pipelines.raphaelmansuy.github.io · read 2 Oct 2026
- Tracing modes
- The site lists convert, logo, premium, auto, smart, and optimal modes for different image types and quality requirements.raphaelmansuy.github.io · read 3 Oct 2026
