Comparison · LightConv vs Squoosh

LightConv vs Squoosh

Google's open-source in-browser image compression playground, well-known for popularising WebAssembly image codecs and side-by-side quality comparison.

This page lays out what Squoosh (founded 2018, free, open-source, no business model. maintained on google's 20% time until 2023 layoffs slowed the project.) does that LightConv doesn't, and what LightConv does that Squoosh doesn't. Both tools exist for a reason — the right pick depends on what you're converting and where the file should be allowed to go.

Squoosh's primary audience: Designers and developers compressing single images for web delivery, comparing visual quality between encoders, exploring codec tradeoffs.

Side-by-side

Dimension
Squoosh
LightConv
Edge
Where conversion happens
In browser (WASM)
In browser (WASM)
Tie
Open source
Yes — Apache 2 on GitHub
No
Squoosh
Side-by-side quality preview
Yes — Squoosh's killer feature
No
Squoosh
Batch conversion
Single file only
Up to 100 files in parallel
LightConv
HEIC support
Dropped in 2022
Yes via libheif WASM
LightConv
Non-image tools
None — images only by design
CSV, JSON, YAML, JWT, Base64, UUID, SHA-256, timestamp
LightConv
Active maintenance
Slowed since Google's 2023 layoffs
Active
LightConv
Per-codec parameter tuning
Extensive (every encoder option exposed)
Quality slider + format only — keeps the UI simple
Squoosh

Squoosh is the better pick if

  • You're tuning quality of ONE image and want to compare side-by-side at different encoder settings
  • You want to inspect or fork the source code — it's Apache-2 on GitHub
  • You only need image compression (Squoosh is single-purpose by design)
  • You're learning about image codecs — Squoosh is the best educational tool for this

LightConv is the better pick if

  • You need to convert HEIC files — Squoosh decoded HEIC for a while then dropped support
  • You're converting more than one file at a time (Squoosh is one-at-a-time)
  • You also need CSV/JSON/YAML/JWT/Base64 — Squoosh is image-only
  • You want a tool that's still actively maintained

Squoosh's strengths — credit where due

  • Side-by-side quality comparison UI with a draggable split-screen — the killer feature LightConv doesn't replicate
  • Per-codec parameter tuning (mozjpeg quality, libavif speed, oxipng level) exposed in the UI for power users
  • Open source (Apache 2) — auditable, forkable, no vendor risk
  • Designed by Google's CDS team — high polish, clean type system

Where LightConv has the structural edge

  • Batch conversion of up to 100 files at once (Squoosh is single-image)
  • More input formats — HEIC, TIFF, BMP, ICO that Squoosh either rejects or doesn't decode
  • Includes non-image tools (CSV, JSON, YAML, JWT, Base64) Squoosh doesn't cover by design
  • Active maintenance — Squoosh's last meaningful commit was over a year ago
  • Programmatic-SEO landing pages for every format pair (32 of them) so Google indexes the specific conversion you need

If LightConv sounds like the right pick for your use case, start at the homepage — every tool is on one page and the conversion happens before you finish reading the card.