diff --git a/src/lib/chapterDetection.ts b/src/lib/chapterDetection.ts
new file mode 100644
index 0000000..4c8c5bf
--- /dev/null
+++ b/src/lib/chapterDetection.ts
@@ -0,0 +1,142 @@
+/**
+ * Utility functions for detecting and splitting text into chapters
+ */
+
+export interface Chapter {
+  title: string;
+  startIndex: number;
+  endIndex: number;
+  text: string;
+}
+
+/**
+ * Detects chapters in text content
+ * Looks for common chapter markers like "Chapter 1", "CHAPTER I", etc.
+ */
+export function detectChapters(text: string, maxChunkSize = 50000): Chapter[] {
+  const lines = text.split('\n');
+  const chapters: Chapter[] = [];
+
+  // Common chapter markers
+  const chapterPatterns = [
+    /^chapter\s+(\d+|[ivxlcdm]+)/i,
+    /^ch\.\s+(\d+)/i,
+    /^(\d+)\.\s+/,  // "1. Title"
+    /^part\s+(\d+|[ivxlcdm]+)/i,
+    /^section\s+(\d+)/i,
+  ];
+
+  let currentChapterStart = 0;
+  let currentChapterTitle = 'Beginning';
+
+  lines.forEach((line, index) => {
+    const trimmed = line.trim();
+
+    // Check if this line is a chapter marker
+    const isChapterMarker = chapterPatterns.some(pattern => pattern.test(trimmed));
+
+    if (isChapterMarker && index > 0) {
+      // Save the previous chapter
+      const chapterText = lines.slice(currentChapterStart, index).join('\n');
+      if (chapterText.trim().length > 0) {
+        chapters.push({
+          title: currentChapterTitle,
+          startIndex: currentChapterStart,
+          endIndex: index,
+          text: chapterText,
+        });
+      }
+
+      // Start new chapter
+      currentChapterStart = index;
+      currentChapterTitle = trimmed || `Chapter ${chapters.length + 1}`;
+    }
+  });
+
+  // Add the final chapter
+  const finalChapterText = lines.slice(currentChapterStart).join('\n');
+  if (finalChapterText.trim().length > 0) {
+    chapters.push({
+      title: currentChapterTitle,
+      startIndex: currentChapterStart,
+      endIndex: lines.length,
+      text: finalChapterText,
+    });
+  }
+
+  // If no chapters were detected, split by size
+  if (chapters.length === 0 || (chapters.length === 1 && text.length > maxChunkSize)) {
+    return splitBySize(text, maxChunkSize);
+  }
+
+  // If chapters are still too large, split them further
+  const refinedChapters: Chapter[] = [];
+  chapters.forEach(chapter => {
+    if (chapter.text.length > maxChunkSize) {
+      const subChapters = splitBySize(chapter.text, maxChunkSize);
+      subChapters.forEach((subChapter, i) => {
+        refinedChapters.push({
+          ...subChapter,
+          title: `${chapter.title} (Part ${i + 1})`,
+        });
+      });
+    } else {
+      refinedChapters.push(chapter);
+    }
+  });
+
+  return refinedChapters.length > 0 ? refinedChapters : chapters;
+}
+
+/**
+ * Splits text into chunks of approximately equal size
+ * Tries to split at paragraph boundaries
+ */
+function splitBySize(text: string, maxChunkSize: number): Chapter[] {
+  const chapters: Chapter[] = [];
+  const paragraphs = text.split(/\n\n+/);
+
+  let currentChunk = '';
+  let chunkIndex = 0;
+
+  paragraphs.forEach((paragraph, index) => {
+    const potentialChunk = currentChunk + (currentChunk ? '\n\n' : '') + paragraph;
+
+    if (potentialChunk.length > maxChunkSize && currentChunk.length > 0) {
+      // Save current chunk
+      chapters.push({
+        title: `Section ${chunkIndex + 1}`,
+        startIndex: 0, // We'll recalculate these if needed
+        endIndex: 0,
+        text: currentChunk,
+      });
+
+      // Start new chunk
+      currentChunk = paragraph;
+      chunkIndex++;
+    } else {
+      currentChunk = potentialChunk;
+    }
+  });
+
+  // Add the final chunk
+  if (currentChunk.trim().length > 0) {
+    chapters.push({
+      title: `Section ${chunkIndex + 1}`,
+      startIndex: 0,
+      endIndex: 0,
+      text: currentChunk,
+    });
+  }
+
+  return chapters;
+}
+
+/**
+ * Estimate the number of chapters that would be created
+ * Useful for showing chapter count before processing
+ */
+export function estimateChapterCount(text: string, maxChunkSize = 50000): number {
+  const chapters = detectChapters(text, maxChunkSize);
+  return chapters.length;
+}