Benchmarks

Each tool fires the same event 1000 times a frame, with the same data, for ten seconds, and every tool received everything it sent. The numbers below are medians. The percentiles, and the harness itself, are in benchmark/Benchmarks.md.

The run was made on 2026-09-24 on an Apple M1 with 16 GB of memory, in Roblox Studio. The versions were BlinkBlox 0.29.0, zap 0.6.29 and ByteNet 0.4.3.

Booleans

The payload is an array of 1000 true values: boolean[0..1000].

Tool	FPS	Kbps
Roblox remotes	15	136800
BlinkBlox	57	3.19
zap	37	8.53
ByteNet	22	8.33

BlinkBlox sends about 2.7 times less data than zap or ByteNet here, because boolean[] is packed eight elements to a byte. The other two spend a whole byte on each boolean.

Entities

The payload is 100 structs of six u8 fields each: an id, a position, an orientation and an animation.

Tool	FPS	Kbps
Roblox remotes	16	721337
BlinkBlox	60	41.57
zap	42	41.86
ByteNet	24	41.71

Every tool encodes six u8 fields in six bytes, so the bandwidth is the same within noise. The frame rate is where they differ: the generated serialisers cost less per event.

Reading the numbers

60 FPS is Studio's cap

Studio does not render faster than 60 frames a second. A tool at 60 FPS is limited by the cap, not by its own cost, so the Entities result shows only that BlinkBlox kept up. It does not show by how much.

• The inbound limits are raised for the run. benchmark/definitions/Definition.blink sets the following options far above their defaults:

  • MaxPacketSize and MaxEventsPerPacket;
  • InboundBytesPerSecond and InboundBurst.

  A thousand events a frame is a flood, and with the defaults a live server refuses most of it. Refusing it is the purpose of those limits. The checks still run, so their cost is included in the numbers above. See Securing the server.

• The events go from client to server. That direction is the one BlinkBlox validates. Every event is checked before a listener sees it.

• Bandwidth depends on your schema. These two payloads are simple. Where a schema has bounded lengths, optionals or booleans spread across structs, the savings come from the techniques in Bandwidth and sizes.

• A different machine gives different numbers. The ratios between the tools are the useful part.

Running them yourself

The harness lives in benchmark/. For now it is Windows-only: it uses 7-Zip, wmic and build.bat to fetch the tools and to record the machine’s specs. On other systems, fetch the tools by hand and fill in the specs yourself.