diff --git a/CHANGELOG.md b/CHANGELOG.md
index 4826f55..0cd7aae 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -23,6 +23,7 @@ Additions
 
 Improvements
 - Stopped the text formatter throwing if a color can't be parsed
+- Improved text formatter tokenization performance
 
 Fixes
 - Fixed TextInput not working correctly when using surrogate pairs
diff --git a/MLEM/Formatting/Codes/Code.cs b/MLEM/Formatting/Codes/Code.cs
index 8b0f75e..393b149 100644
--- a/MLEM/Formatting/Codes/Code.cs
+++ b/MLEM/Formatting/Codes/Code.cs
@@ -23,9 +23,9 @@ namespace MLEM.Formatting.Codes {
         public readonly Match Match;
         /// <summary>
         /// The tokens that this formatting code is a part of.
-        /// Note that this array only has multiple entries if additional tokens have to be started while this code is still applied.
+        /// Note that this collection only has multiple entries if additional tokens have to be started while this code is still applied.
         /// </summary>
-        public IList<Token> Tokens { get; internal set; }
+        public readonly List<Token> Tokens = new List<Token>();
 
         /// <summary>
         /// Creates a new formatting code based on a formatting code regex and its match.
diff --git a/MLEM/Formatting/TextFormatter.cs b/MLEM/Formatting/TextFormatter.cs
index 1407572..f17124d 100644
--- a/MLEM/Formatting/TextFormatter.cs
+++ b/MLEM/Formatting/TextFormatter.cs
@@ -156,11 +156,12 @@ namespace MLEM.Formatting {
             // resolve macros
             s = this.ResolveMacros(s);
             var tokens = new List<Token>();
-            var codes = new List<Code>();
+            var applied = new List<Code>();
+            var allCodes = new List<Code>();
             // add the formatting code right at the start of the string
             var firstCode = this.GetNextCode(s, 0, 0);
             if (firstCode != null)
-                codes.Add(firstCode);
+                applied.Add(firstCode);
             var index = 0;
             var rawIndex = 0;
             while (rawIndex < s.Length) {
@@ -168,24 +169,25 @@ namespace MLEM.Formatting {
                 // if we've reached the end of the string
                 if (next == null) {
                     var sub = s.Substring(rawIndex, s.Length - rawIndex);
-                    tokens.Add(new Token(codes.ToArray(), index, rawIndex, TextFormatter.StripFormatting(font, sub, codes), sub));
+                    tokens.Add(new Token(applied.ToArray(), index, rawIndex, TextFormatter.StripFormatting(font, sub, applied), sub));
                     break;
                 }
+                allCodes.Add(next);
 
                 // create a new token for the content up to the next code
                 var ret = s.Substring(rawIndex, next.Match.Index - rawIndex);
-                var strippedRet = TextFormatter.StripFormatting(font, ret, codes);
-                tokens.Add(new Token(codes.ToArray(), index, rawIndex, strippedRet, ret));
+                var strippedRet = TextFormatter.StripFormatting(font, ret, applied);
+                tokens.Add(new Token(applied.ToArray(), index, rawIndex, strippedRet, ret));
 
                 // move to the start of the next code
                 rawIndex = next.Match.Index;
                 index += strippedRet.Length;
 
                 // remove all codes that are incompatible with the next one and apply it
-                codes.RemoveAll(c => c.EndsHere(next) || next.EndsOther(c));
-                codes.Add(next);
+                applied.RemoveAll(c => c.EndsHere(next) || next.EndsOther(c));
+                applied.Add(next);
             }
-            return new TokenizedString(font, alignment, s, TextFormatter.StripFormatting(font, s, tokens.SelectMany(t => t.AppliedCodes)), tokens.ToArray());
+            return new TokenizedString(font, alignment, s, TextFormatter.StripFormatting(font, s, allCodes), tokens.ToArray(), allCodes.ToArray());
         }
 
         /// <summary>
diff --git a/MLEM/Formatting/Token.cs b/MLEM/Formatting/Token.cs
index 029c8ca..6280351 100644
--- a/MLEM/Formatting/Token.cs
+++ b/MLEM/Formatting/Token.cs
@@ -50,6 +50,8 @@ namespace MLEM.Formatting {
             this.RawIndex = rawIndex;
             this.Substring = substring;
             this.RawSubstring = rawSubstring;
+            foreach (var code in appliedCodes)
+                code.Tokens.Add(this);
         }
 
         /// <summary>
diff --git a/MLEM/Formatting/TokenizedString.cs b/MLEM/Formatting/TokenizedString.cs
index db49998..f27317b 100644
--- a/MLEM/Formatting/TokenizedString.cs
+++ b/MLEM/Formatting/TokenizedString.cs
@@ -1,11 +1,9 @@
 using System;
 using System.Collections.Generic;
-using System.Collections.ObjectModel;
 using System.Linq;
 using System.Text;
 using Microsoft.Xna.Framework;
 using Microsoft.Xna.Framework.Graphics;
-using MLEM.Extensions;
 using MLEM.Font;
 using MLEM.Formatting.Codes;
 using MLEM.Misc;
@@ -42,17 +40,11 @@ namespace MLEM.Formatting {
         private float initialInnerOffset;
         private RectangleF area;
 
-        internal TokenizedString(GenericFont font, TextAlignment alignment, string rawString, string strg, Token[] tokens) {
+        internal TokenizedString(GenericFont font, TextAlignment alignment, string rawString, string strg, Token[] tokens, Code[] allCodes) {
             this.RawString = rawString;
             this.String = strg;
             this.Tokens = tokens;
-
-            // since a code can be present in multiple tokens, we use Distinct here
-            this.AllCodes = tokens.SelectMany(t => t.AppliedCodes).Distinct().ToArray();
-            // TODO this can probably be optimized by keeping track of a code's tokens while tokenizing
-            foreach (var code in this.AllCodes)
-                code.Tokens = new ReadOnlyCollection<Token>(this.Tokens.Where(t => t.AppliedCodes.Contains(code)).ToList());
-
+            this.AllCodes = allCodes;
             this.Realign(font, alignment);
         }