diff --git a/src/cm2py/memory/__init__.py b/src/cm2py/memory/__init__.py
new file mode 100644
index 0000000..b63e293
--- /dev/null
+++ b/src/cm2py/memory/__init__.py
@@ -0,0 +1,3 @@
+#!/usr/bin/env python3
+
+from .memory import *
diff --git a/src/cm2py/memory/memory.py b/src/cm2py/memory/memory.py
new file mode 100644
index 0000000..2af934f
--- /dev/null
+++ b/src/cm2py/memory/memory.py
@@ -0,0 +1,276 @@
+from string import ascii_lowercase, ascii_uppercase, digits
+from typing import Union, Callable, IO
+import logging, base64, math, struct
+
+try:
+    import zlib
+except:
+    zlib = None
+    logging.exception("'zlib' wasn't found. Install it to use HugeMemory.")
+
+class Memory:
+    """Base memory class; any memory. This class uses a list of integers (rather than bytes) as means of storing/processing data."""
+    raw = []
+    """The raw data in the memory. It is not recommended to use this directly."""
+    size = 0
+    """Size of the memory (in bytes). This number is the number of addresses, not the total data."""
+    nbits = 0
+    """Bits of data at each address."""
+    __readable = 0
+    """Readability; Not Implemented"""
+    __writable = 0
+    """Writability; Not Implemented"""
+
+    __encoder = None
+    """The encoding function to use; data -> memstring"""
+    __decoder = None
+    """The decoding function to use; memstring -> data"""
+
+    def __init__(self, size : int, nbits : int, encoder : Callable, decoder : Callable, init_data = None):
+        """You shouldn't use this class unless you're making your own memory type. Use MassMemory, MassiveMemory, etc instead.
+        :param int size: The size (number of addresses)
+        :param int nbits: The number of bits per address (must be a multiple of 8, and non-zero)
+        :param encoder: An encoding function to encode data into memstrings. See massMemoryEncode as an example
+        :param decoder: A decoder function to get data from a memstring. See massMemoryDecode as an example
+        :param init_data: Optional initial data. If it's a memstring, it's decoded using the provided decoder, otherwise it's assumed to just be raw data."""
+        self.size = size
+        self.nbits = nbits
+        self.__encoder = encoder
+        self.__decoder = decoder
+        if init_data:
+            if isinstance(init_data, str):
+                self.load(init_data)
+            else:
+                self.raw = init_data
+        if len(self.raw) < self.size:
+            self.raw.extend([0] * (self.size - len(self.raw)))
+        
+
+    def get(self, index : int) -> int:
+        """Get data at an index. Alternatively, use Memory[index]. Negative indexes allowed."""
+        if index >= self.size:
+            raise IndexError(f"{index} is greater than maximum of {self.size-1}.")
+        return self.raw[index]
+
+    def put(self, data : int, index: int) -> None:
+        """Put data at an index. Alternatively, use Memory[index] = data. Negative indexes allowed."""
+        if index >= self.size:
+            raise IndexError(f"{index} is greater than maximum of {self.size-1}.")
+        self.raw[index] = data
+
+    def load(self, text : str) -> None:
+        """Load a memory string into this memory (overwrites existing contents)."""
+        if not self.__decoder:
+            raise NotImplementedError("A decoding function was not specified.")
+        self.raw = self.__decoder(text)
+
+    def save(self) -> str:
+        """Save memory contents as a memstring."""
+        if not self.__encoder:
+            raise NotImplementedError("An encoding function was not specified.")
+        return self.__encoder(self.raw)
+    
+    def dumpfile(self, file : Union[str, IO[bytes]], n:int=-1) -> None:
+        """Dump memory data to a file or buffer
+        :param int n: Max number of addresses to write. -1 for infinite
+        """
+        if n < 0:
+            n = 2**32
+        i=0
+        if isinstance(file, str):
+            with open(file, 'wb') as f:
+                for v in self.raw:
+                    f.write(v.to_bytes(math.ceil(self.nbits / 8), "little"))
+                    i += 1
+                    if i >= n:
+                        break
+        else:
+            if file.seekable():
+                file.seek(0)
+            for v in self.raw:
+                file.write(v.to_bytes(math.ceil(self.nbits / 8), "little"))
+                i+=1
+                if i >= n:
+                    break
+            
+
+    def loadfile(self, file : Union[str, IO[bytes]]) -> None:
+        """Load memory data from a file or buffer. Contents are truncated to fit"""
+        if isinstance(file, str):
+            with open(file, 'rb') as f:
+                for i in range(self.size):
+                    self.raw[i] = int.from_bytes(f.read(math.ceil(self.nbits / 8)), "little")
+        else:
+            if file.seekable():
+                file.seek(0)
+            for i in range(self.size):
+                    self.raw[i] = int.from_bytes(file.read(math.ceil(self.nbits / 8)), "little")
+        if len(self.raw) < self.size:
+            self.raw.extend([0] * (self.size - len(self.raw)))
+
+    def __getitem__(self, key : Union[str, int]) -> int:
+        """Integers and bin strings can be used interchangeably for the index."""
+        if isinstance(key,str):
+            if key.startswith('0b'):
+                key = key[2:]
+            key = int(key,2)
+        return self.raw[key]
+    def __setitem__(self, key : Union[str, int], value : Union[str, int]) -> None:
+        """Integers and bin strings can be used for both parameters interchangeably."""
+        if isinstance(key,str):
+            if key.startswith('0b'):
+                key = key[2:]
+            key = int(key,2)
+        if isinstance(value,str):
+            if value.startswith('0b'):
+                value = value[2:]
+            value = int(value,2)
+        self.raw[key] = value
+
+def load(data : str) -> Memory:
+    """Load a memory string. Automatically detects memory type. Will raise an exception for invalid memstrings.
+    If you already know what type of memory the string belongs to, you should use MassMemory(memstring), MassiveMemory(memstring), etc"""
+    if len(data) == 8192:
+        target = MassMemory
+    elif len(data) == 12288:
+        target = MassiveMemory
+    else:
+        target = HugeMemory
+    return target(data)
+
+def massMemoryEncode(data: list[int]) -> str:
+    """Encodes data into a MassMemory memstring."""
+    return ''.join(f'{i:02X}' for i in data)
+
+def massMemoryDecode(text: str) -> list[int]:
+    """Decodes a MassMemory memstring into data."""
+    return [int(text[i:i+2], 16) for i in range(0, len(text), 2)]
+
+def massiveMemoryEncode(data: list[int]) -> str:
+    """Encodes data into a MassiveMemory memstring."""
+    alphabet = ascii_uppercase + ascii_lowercase + digits + "+/"
+    hex_str = ''.join(f'{i:04x}' for i in data)
+    instruction_bytes = bytes.fromhex(hex_str)
+
+    output_chars = []
+    for b1, b2 in zip(instruction_bytes[::2], instruction_bytes[1::2]):
+        bits = (b1 << 8) | b2
+        output_chars.extend(alphabet[(bits >> (6 * i)) & 0x3F] for i in range(3))
+    
+    output = ''.join(output_chars)
+    return output.ljust(12288, 'A')
+
+def massiveMemoryDecode(text: str) -> list[int]:
+    """Decodes a MassiveMemory memstring into data."""
+    alphabet = ascii_uppercase + ascii_lowercase + digits + "+/"
+    index_table = {c: i for i, c in enumerate(alphabet)}
+
+    outb = []
+    for i in range(0, len(text), 3):
+        group = text[i:i + 3]
+        bits = 0
+        for c in group:
+            bits = (bits << 6) | index_table[c]
+        outb.append(bits)
+    return outb
+
+def hugeMemoryEncode(data: list[int]) -> str:
+    """Encodes data into a HugeMemory memstring."""
+    packed = struct.pack(f'<{len(data)}H', *data)  # Little-endian 16-bit unsigned ints
+    compressed = zlib.compress(packed)[2:-4]       # Strip zlib headers and checksum (raw deflate)
+    return base64.b64encode(compressed).decode('ascii')
+
+def hugeMemoryDecode(text: str) -> list[int]:
+    """Decodes a HugeMemory memstring into data."""
+    rs = zlib.decompress(base64.b64decode(text), wbits=-zlib.MAX_WBITS)
+    return list(struct.unpack(f'<{len(rs)//2}H', rs))
+
+class MassMemory(Memory):
+    """MassMemory; 8 bit addresses with 8 bit data."""
+    size = 2**8
+
+    def __init__(self, data : Union[str, list[int], None]=None):
+        """
+        :param data: Optional initial data. A memstring can be provided, or just data. 
+        """
+        super().__init__(2**8, 8, massMemoryEncode, massMemoryDecode, init_data=data)
+
+class MassiveMemory(Memory):
+    """MassiveMemory; 12 bit addresses with 16 bit data."""
+    size = 2**12
+    def __init__(self, data : Union[str, list[int], None]=None):
+        """
+        :param data: Optional initial data. A memstring can be provided, or just data. 
+        """
+        super().__init__(2**12, 16, massiveMemoryEncode, massiveMemoryDecode, init_data=data)
+
+class HugeMemory(Memory):
+    """HugeMemory; 16 bit addresses with 16 bit data."""
+    size = 2**16
+    def __init__(self, data : Union[str, list[int], None]=None):
+        """
+        :param data: Optional initial data. A memstring can be provided, or just data. 
+        """
+        super().__init__(2**16, 16, hugeMemoryEncode, hugeMemoryDecode, init_data=data)
+
+class ROMs:
+    """Pre-defined ROM loaders. These load a pre-defined ROM onto an existing Memory."""
+    def divider(mem_instance : Memory) -> None:
+        """6-bit divider (0 / 0 = 0). Outputs are rounded."""
+        if not mem_instance.massive:
+            raise Exception("Memory is not a MassiveMemory")
+        fp = mem_instance
+        for i in range(4096):
+            param1 = (i & 0b111111000000)>>6
+            param2 = (i & 0b000000111111)
+            if param2 == 0:
+                fp[i] = 0
+            else:
+                fp[i] = int(round(param1/param2))
+
+    def multiplier(mem_instance : Memory) -> None:
+        """6-bit multiplier."""
+        if not mem_instance.massive:
+            raise Exception("Memory is not a MassiveMemory")
+        fp = mem_instance
+        for i in range(4096):
+            param1 = (i & 0b111111000000)>>6
+            param2 = (i & 0b000000111111)
+            fp[i] = param1*param2
+
+    def adder(mem_instance : Memory) -> None:
+        """6-bit adder."""
+        if not mem_instance.massive:
+            raise Exception("Memory is not a MassiveMemory")
+        fp = mem_instance
+        for i in range(4096):
+            param1 = (i & 0b111111000000)>>6
+            param2 = (i & 0b000000111111)
+            fp[i] = param1+param2
+        
+    def subtractor(mem_instance : Memory) -> None:
+        """6-bit subtractor (any result below 0 just outputs 0)."""
+        if not mem_instance.massive:
+            raise Exception("Memory is not a MassiveMemory")
+        fp = mem_instance
+        for i in range(4096):
+            param1 = (i & 0b111111000000)>>6
+            param2 = (i & 0b000000111111)
+            if param1-param2 < 0:
+                fp[i] = 0
+            else:
+                fp[i] = param1-param2
+
+    def doubledabble(mem_instance: Memory) -> None:
+        """Outputs the input number's digits as 4 4-bit numbers. Example: 1024 -> 4 2 0 1"""
+        if not mem_instance.massive:
+            raise Exception("Memory is not a MassiveMemory")
+        fp = mem_instance
+        for i in range(4096):
+            st = f'{i:04}'  # 4-character decimal string, zero-padded
+            fp[i] = (
+                f'{int(st[3]):04b}' +
+                f'{int(st[2]):04b}' +
+                f'{int(st[1]):04b}' +
+                f'{int(st[0]):04b}'
+            )
diff --git a/src/cm2py/utilities/utilities.py b/src/cm2py/utilities/utilities.py
index bfc8265..5862ad4 100644
--- a/src/cm2py/utilities/utilities.py
+++ b/src/cm2py/utilities/utilities.py
@@ -172,40 +172,6 @@ def generateDecoder(
 
 base64 = string.ascii_uppercase + string.ascii_lowercase + string.digits + "+/"
 
-
-def encodeToMemory(
-    data: list[int], memoryType: Literal["mass", "massive", "huge"]
-) -> str:
-    """
-    Turns a list of integers into a string that can be pasted into one of the memory buildings.
-    """
-
-    assert memoryType in [
-        "mass",
-        "massive",
-        "huge",
-    ], 'Invalid memory building type. Use "mass", "massive", or "huge"'
-
-    code = ""
-
-    if memoryType == "mass":
-        for v in data:
-            code += format(v % 256, "02x")
-        code += "00" * (4096 - len(data))
-    elif memoryType == "massive":
-        for v in data:
-            code += base64[v & 0x3F]
-            code += base64[(v >> 6) & 0x3F]
-            code += base64[(v >> 12) & 0x3F]
-        code += "AAA" * (4096 - len(data))
-    elif memoryType == "huge":
-        raise (
-            NotImplementedError,
-            "Huge Memory uses full utf8 to represent the values, which don't work well with Roblox yet. When the format gets updated or full utf8 is supported, this function will be updated.",
-        )
-    return code
-
-
 def halfPrecisionBitsToNumber(bits: int) -> float:
     """
     Converts a half-precision floating point number stored in an integer to a python float.