diff --git a/services/web/frontend/js/features/source-editor/lezer-typst/tokens.mjs b/services/web/frontend/js/features/source-editor/lezer-typst/tokens.mjs
index a29d433106..d9ca805889 100644
--- a/services/web/frontend/js/features/source-editor/lezer-typst/tokens.mjs
+++ b/services/web/frontend/js/features/source-editor/lezer-typst/tokens.mjs
@@ -252,19 +252,36 @@ export const lineCommentContentTokenizer = new ExternalTokenizer(
 )
 
 // ── mathContentTokenizer ────────────────────────────────────────────────
-// Emits MathContent — everything between the $...$ delimiters (no newlines).
-// External rather than a @tokens rule for the same reason as LineCommentContent:
-// ![$\n]+ overlaps with spaces, '<', '@', and other literals in merged states.
+// Emits MathContent — everything between the $...$ delimiters.
+//
+// Typst distinguishes inline math ($x^2$) from display math ($ x^2 $):
+// display math has whitespace between the opening '$' and the content.
+// We detect this by scanning back to '$': if there is any whitespace
+// between '$' and the current position (i.e. @skip consumed it), the
+// tokenizer allows newlines so multi-line display math works.  Inline math
+// keeps the newline stop, preventing a lone '$' from consuming the rest of
+// the document.
+//
+// contextual: true — only fires inside InlineMath after '$', never in
+// body text.  The '$' token appears nowhere else in the grammar so the
+// post-'$' state does not merge with item* states.
 export const mathContentTokenizer = new ExternalTokenizer(
   (input, _stack) => {
+    // Scan back to the opening '$', detecting display vs inline math.
+    let back = -1
+    while (input.peek(back) === SPACE || input.peek(back) === TAB || input.peek(back) === NEWLINE) back--
+    if (input.peek(back) !== DOLLAR) return
+    const isDisplay = back < -1  // whitespace between '$' and current position
+
     let hasContent = false
-    while (input.next !== -1 && input.next !== DOLLAR && input.next !== NEWLINE) {
+    while (input.next !== -1 && input.next !== DOLLAR) {
+      if (!isDisplay && input.next === NEWLINE) break
       input.advance()
       hasContent = true
     }
     if (hasContent) input.acceptToken(MathContent)
   },
-  { contextual: false }
+  { contextual: true }
 )
 
 // ── codeKeywordTokenizer ─────────────────────────────────────────────────