diff --git a/src/PacketReader.h b/src/PacketReader.h
index de71de3..c6912f7 100644
--- a/src/PacketReader.h
+++ b/src/PacketReader.h
@@ -1,6 +1,7 @@
 #pragma once
 
 #include "Common.h"
+#include "util/ByteUtil.h"
 
 #include <cstdio>
 #include <cstdint>
@@ -19,20 +20,32 @@ namespace Feather
         {
         }
 
-        template <typename T>
+        // Peek the next value to be read
+        // If big endian (the default), will reverse byte order.
+        template <typename T, Endian endianMode = Endian::Big>
         inline T Peek()
         {
-            return *reinterpret_cast<const T* const>(&m_data[m_offset]);
+            T value = *reinterpret_cast<const T *const>(&m_data[m_offset]);
+
+            return (endianMode == Endian::Big) ? ReverseBytes<T>(value) : value;
         }
 
-        template <typename T>
+        // Read the next value
+        // If big endian (the default), will reverse byte order.
+        template <typename T, Endian endianMode = Endian::Big>
         inline T Read()
         {
-            T value = Peek<T>();
+            T value = Peek<T, endianMode>();
             m_offset += sizeof(T);
             return value;
         }
 
+        // Fast way to peek the next byte
+        inline uint8_t PeekByte() { return m_data[m_offset]; }
+
+        // Fast way to read a byte
+        inline uint8_t ReadByte() { return m_data[m_offset++]; }
+
         template <typename T = int32_t>
         inline T ReadVarInt()
         {
@@ -41,7 +54,7 @@ namespace Feather
             uint8_t read;
             do
             {
-                read = Read<uint8_t>();
+                read = ReadByte();
                 int value = (read & 0b01111111);
                 result |= (value << (7 * numRead));
 
@@ -62,10 +75,8 @@ namespace Feather
             string str;
 
             for (int i = 0; i < size; i++) {
-                str += Read<uint8_t>();
+                str += ReadByte();
             }
-
-            //printf("Read string of length %d: %s\n", size, str);
             
             return str;
         }
@@ -77,4 +88,11 @@ namespace Feather
         const uint32_t       m_length;
     };
 
+    // Use fast Read and Peek for uint8_t
+
+    template <>
+    inline uint8_t PacketReader::Peek<uint8_t>() { return PeekByte(); }
+
+    template<>
+    inline uint8_t PacketReader::Read<uint8_t>() { return ReadByte(); }
 }
diff --git a/src/util/ByteUtil.h b/src/util/ByteUtil.h
new file mode 100644
index 0000000..eda6fb0
--- /dev/null
+++ b/src/util/ByteUtil.h
@@ -0,0 +1,38 @@
+#pragma once
+
+#ifdef _MSC_VER
+#include <stdlib.h>
+#endif
+
+enum class Endian
+{
+    Little,
+    Big
+};
+
+// Reverses the byte order of a primitive value of any typical size
+template <typename T>
+inline constexpr T ReverseBytes(T n)
+{
+    // 1 byte, cannot reverse
+    if constexpr (sizeof(T) == 1) return n;
+
+#ifdef __GNUC__
+    // GCC intrinsic byte swaps
+    // https://gcc.gnu.org/onlinedocs/gcc/Other-Builtins.html#index-_005f_005fbuiltin_005fbswap16
+    
+    if constexpr (sizeof(T) == 2)      return __builtin_bswap16(n); // 2 bytes
+    else if constexpr (sizeof(T) == 4) return __builtin_bswap32(n); // 4 bytes
+    else if constexpr (sizeof(T) == 8) return __builtin_bswap64(n); // 8 bytes
+
+#elif defined(_MSC_VER)
+    // MSVC intrinsic byteswaps
+    // https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/byteswap-uint64-byteswap-ulong-byteswap-ushort
+
+    if constexpr (sizeof(T) == 2)      return _byteswap_ushort(n); // 2 bytes
+    else if constexpr (sizeof(T) == 4) return _byteswap_ulong(n);  // 4 bytes
+    else if constexpr (sizeof(T) == 8) return _byteswap_uint64(n); // 8 bytes
+#endif
+    //else static_assert(false, "Attempted to call ReverseBytes() on type with unusual size.");
+}
+