diff --git a/middleware.ts b/middleware.ts
new file mode 100644
index 000000000..a853abbc9
--- /dev/null
+++ b/middleware.ts
@@ -0,0 +1,55 @@
+import { NextRequest, NextResponse } from "next/server";
+
+/**
+ * Middleware that handles content negotiation for markdown requests.
+ * When a request includes Accept: text/markdown header, it rewrites
+ * the request to the markdown API endpoint to return raw markdown content.
+ */
+export function middleware(request: NextRequest) {
+  const acceptHeader = request.headers.get("accept") || "";
+
+  // Check if the client is requesting markdown content
+  const wantsMarkdown =
+    acceptHeader.includes("text/markdown") ||
+    acceptHeader.includes("text/x-markdown");
+
+  if (!wantsMarkdown) {
+    return NextResponse.next();
+  }
+
+  const pathname = request.nextUrl.pathname;
+
+  // Skip API routes, static files, and special paths
+  if (
+    pathname.startsWith("/api/") ||
+    pathname.startsWith("/_next/") ||
+    pathname.startsWith("/images/") ||
+    pathname.includes(".") // Skip files with extensions (e.g., .png, .js, .css)
+  ) {
+    return NextResponse.next();
+  }
+
+  // Skip the root path
+  if (pathname === "/") {
+    return NextResponse.next();
+  }
+
+  // Rewrite to the markdown API endpoint
+  const url = request.nextUrl.clone();
+  url.pathname = `/api/markdown${pathname}`;
+
+  return NextResponse.rewrite(url);
+}
+
+export const config = {
+  // Match all paths except static files and API routes
+  matcher: [
+    /*
+     * Match all request paths except for:
+     * - _next/static (static files)
+     * - _next/image (image optimization files)
+     * - favicon.ico (favicon file)
+     */
+    "/((?!_next/static|_next/image|favicon.ico).*)",
+  ],
+};
diff --git a/pages/api/markdown/[...slug].ts b/pages/api/markdown/[...slug].ts
new file mode 100644
index 000000000..8a6a7827b
--- /dev/null
+++ b/pages/api/markdown/[...slug].ts
@@ -0,0 +1,72 @@
+import fs from "fs";
+import { join } from "path";
+import type { NextApiRequest, NextApiResponse } from "next";
+import { CONTENT_DIR, DOCS_FILE_EXTENSIONS } from "@/lib/content.server";
+
+/**
+ * API route that serves raw markdown/MDX content for a given slug.
+ * This enables content negotiation - when a client requests a page with
+ * Accept: text/markdown, they receive the raw markdown instead of HTML.
+ */
+export default function handler(req: NextApiRequest, res: NextApiResponse) {
+  if (req.method !== "GET") {
+    return res
+      .status(405)
+      .setHeader("Allow", "GET")
+      .json({ error: `${req.method} method is not accepted.` });
+  }
+
+  // Get slug from query params, or parse from URL path if not available
+  // (middleware rewrites may not populate req.query properly)
+  let slug: string | string[] | undefined = req.query.slug;
+
+  if (!slug || (Array.isArray(slug) && slug.length === 0)) {
+    // Parse slug from the URL path: /api/markdown/path/to/page -> ["path", "to", "page"]
+    const urlPath = req.url?.split("?")[0] || "";
+    const pathSegments = urlPath.replace(/^\/api\/markdown\/?/, "").split("/");
+    slug = pathSegments.filter((s) => s.length > 0);
+  }
+
+  if (!slug || (Array.isArray(slug) && slug.length === 0)) {
+    return res.status(400).json({ error: "Invalid slug parameter" });
+  }
+
+  // Ensure slug is an array
+  const slugArray = Array.isArray(slug) ? slug : [slug];
+
+  // Try to find the content file with .mdx or .md extension
+  let source: Buffer | null = null;
+  let sourcePath: string | null = null;
+
+  for (const ext of DOCS_FILE_EXTENSIONS) {
+    const candidatePath = join(
+      CONTENT_DIR,
+      ...slugArray.slice(0, slugArray.length - 1),
+      `${slugArray[slugArray.length - 1]}${ext}`,
+    );
+
+    if (fs.existsSync(candidatePath)) {
+      source = fs.readFileSync(candidatePath);
+      sourcePath = candidatePath;
+      break;
+    }
+  }
+
+  // Skip API reference pages
+  if (
+    sourcePath &&
+    (sourcePath.includes("content/__mapi-reference") ||
+      sourcePath.includes("content/__api-reference"))
+  ) {
+    return res.status(404).json({ error: "Page not found" });
+  }
+
+  if (!source) {
+    return res.status(404).json({ error: "Page not found" });
+  }
+
+  // Return the raw markdown content
+  res.setHeader("Content-Type", "text/markdown; charset=utf-8");
+  res.setHeader("Cache-Control", "public, max-age=3600, s-maxage=3600");
+  res.status(200).send(source.toString());
+}