Data types

Data types describe how ByteNet Max serializes values. Primitive exports are values; constructors such as struct, array, map, and optional are functions.

Numbers

Type	Value	Encoded size	Range / precision
uint8	Integer	1 byte	0 to 255
uint16	Integer	2 bytes	0 to 65,535
uint32	Integer	4 bytes	0 to 4,294,967,295
int8	Integer	1 byte	−128 to 127
int16	Integer	2 bytes	−32,768 to 32,767
int32	Integer	4 bytes	−2,147,483,648 to 2,147,483,647
float32	Number	4 bytes	32-bit floating point
float64	Number	8 bytes	64-bit floating point

Score = ByteNetMax.uint32
Temperature = ByteNetMax.float32

Roblox and primitive values

Type	Luau value	Encoded size	Notes
bool	boolean	1 byte	0 or 1
string	string	2-byte length + data	Length prefix is uint16
vec2	Vector2	8 bytes	Two float32 components
vec3	Vector3	12 bytes	Three float32 components
color3	Color3	3 bytes	RGB channels quantized to 8 bits
cframe	CFrame	24 bytes	Position and axis-angle rotation as float32s
buff	buffer	4-byte length + data	Copies the buffer payload
inst	Instance?	2-byte reference index	Uses the transport's reference list
unknown	unknown	2-byte reference index	Uses the transport's reference list
nothing	nil	0 bytes	Useful for empty query requests

struct(fields)

Serializes a fixed dictionary shape.

local PlayerState = ByteNetMax.struct({
    Health = ByteNetMax.uint16,
    Position = ByteNetMax.vec3,
    Name = ByteNetMax.string,
})

The encoded size is the sum of its field sizes.

array(valueType)

Serializes a sequential array. Its count uses a 2-byte uint16 prefix.

local Tags = ByteNetMax.array(ByteNetMax.string)

map(keyType, valueType)

Serializes a dictionary as key/value pairs. Its count uses a 2-byte uint16 prefix.

local Scores = ByteNetMax.map(ByteNetMax.string, ByteNetMax.uint32)

optional(valueType)

Serializes a 1-byte presence marker, followed by the value when it is not nil.

local Subtitle = ByteNetMax.optional(ByteNetMax.string)

auto

Detects the value's type at runtime and adds a 1-byte type marker. Integers use the smallest fitting numeric codec; exact float32 values use float32, otherwise float64 is used.

Supported direct codecs are nil, boolean, number, string, Vector2, Vector3, Color3, and CFrame. Other values use the reference-based unknown codec.

Aliases

Alias	Equivalent type
playerName	string
playerIdentifier	uint8

These are compatibility aliases in the current public module. For Roblox UserId values, prefer a type with sufficient range such as uint32.