// the take
SVGO is the de facto SVG minifier for the JS ecosystem — a plugin-based optimizer that strips redundant editor cruft (metadata, hidden elements, default attribute values) and rewrites path data more compactly. It's for anyone shipping SVGs in a web build: icon libraries, design tool exports, static sites that want smaller asset payloads.
The plugin architecture is the right abstraction — preset-default bundles sane defaults but every plugin can be disabled or reconfigured per-project rather than forcing an all-or-nothing transform. It exposes both a CLI and a plain `optimize()` function with no Node-specific dependencies in the core, which is why it got embedded inside webpack loaders, imagemin, and framework build pipelines rather than staying a standalone tool. 22k stars and this many downstream integrations means the plugin edge cases (weird transform matrices, nested defs, inline styles) have mostly been found already.
convertPathData and the shape-to-path conversions are lossy by nature — they optimize for file size, not pixel-perfect fidelity, so complex or hand-tuned SVGs can render subtly differently after a pass; you need visual regression testing before trusting it in a pipeline, not just a diff of the output markup. The v3→v4 migration doc signals real breaking changes in config format (now ESM-only `.mjs`), so upgrading isn't a drop-in version bump if you have an existing config. It's a single-purpose minifier — no sprite-sheet generation, no icon-font workflow, so you'll still need another tool if that's part of your asset pipeline. And because it mutates/removes attributes based on static analysis of the SVG AST, SVGs that rely on external CSS or JS manipulating attributes SVGO considers 'unused' can break at runtime in ways that only show up in the browser, not in the optimizer's own output.
We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →