Measured - byte-exact - 2026

Pedulli vs brotli

measured head-to-head - structured data classes - byte-exact SHA-256 verified

TL;DR - honest
Pedulli is a best-of-N racer: it races xz, zstd, brotli and your data's SRD math, then keeps the smallest verified output, always byte-exact (SHA-256 roundtrip). Because brotli is one of the coders it runs, the Orchestrator is never larger than the best standard codec (worst case +1 byte). It wins outright on structured classes (JSON, periodic, sparse); on small generic text and already-optimal/random data it simply selects the best codec - a tie at that codec's size, never a loss.

The measured table

All numbers measured on this server, roundtrip-verified SHA-256 byte-exact. Proofs available on request.

Input	brotli -11	Pedulli (best-of-N)	Δ
JSON 31 KB	2,510 B	1,265 B	-50%
Small generic text 4 KB	~1.4 KB	~1.4 KB	tie (selects brotli)
HTML 161 KB	35,705 B	35,705 B	tie (selects brotli)
JavaScript source 47 KB	12,225 B	12,225 B	tie (selects brotli)
1 MiB of zeros	1,036 B	13 B	much smaller (redundant input)
1 MiB periodic binary	1,948 B	436 B	-78%
MP4 master 10 MB	9,830,512 B	9,430,108 B	-400 KB
Apache logs 3.5 MB	231 KB	217 KB	-6.4%
Random bytes 1 MB	1,048,581 B (+5 B)	1,048,577 B (+1 B)	4 B less overhead

Where brotli is the right tool on the wire (honest)

Small generic text. brotli ships a ~122 KB dictionary of common web terms that gives it the edge on short generic English / HTML at level 11 - so the Orchestrator simply selects brotli there and ties its size. Output is never larger.
HTTP Content-Encoding standard. brotli is built into every modern browser. Pedulli is a separate service / API.
Excellent +5 byte random overhead. One of the few codecs that comes close to Pedulli's +1 byte floor.

What Pedulli does that brotli does not

+1 byte never-worse floor. Pedulli's own output is never more than 1 byte over the input - beats even brotli's excellent +5 B.
Best-of-N racing. Several internal coders run; the smallest byte-exact result wins, so structured JSON and domain-shaped data come out tighter.
Web-native + API. In-browser for small inputs, REST API for archival pipelines. brotli on the wire is for HTTP transport.
Zero data retention. GDPR by design, EU-sovereign.

When to switch

Switch to Pedulli for JSON / domain-shaped data archival, video masters, periodic sensor data and the per-file +1 byte never-worse floor - and because, by racing brotli internally, its output is never larger than brotli's. Keep brotli for HTTP Content-Encoding on the wire (browsers already speak it).

Run the free trial Pricing / API All benchmarks

More comparisons: gzip - zstd - xz - 7-Zip - zpaq