GlassBox launches browser fingerprint bench
GlassBox is a single-file, client-side browser fingerprinting bench that runs roughly 31 probes across canvas, WebGL, audio, fonts, WebRTC, permissions, APIs, and more. It exposes raw signals and estimates identifiability while keeping measurements local, with IP geolocation as an opt-out network exception.
GlassBox makes an invisible tracking problem tangible, but its headline identifiability score is best treated as an educational estimate—not proof of real-world uniqueness.
- –Combines weak signals into a practical picture of how trackers and anti-fraud systems link browsers
- –Highlights high-entropy surfaces including canvas, GPU, audio, installed fonts, voices, and browser API support
- –Its static, dependency-free, MIT-licensed design makes the tool easy to audit, fork, and self-host
- –The local-first model protects against the scanner becoming another source of telemetry, though IP intelligence still calls public APIs
- –Community testing exposed limitations around anti-fingerprinting, browser farbling, and imperfect hardware/network readings
DISCOVERED
46d ago
2026-08-25
PUBLISHED
46d ago
2026-08-24
RELEVANCE
AUTHOR
tke248