diff --git a/Userland/Libraries/LibJS/Runtime/Temporal/Duration.cpp b/Userland/Libraries/LibJS/Runtime/Temporal/Duration.cpp
index fde758ccff..b531f444c8 100644
--- a/Userland/Libraries/LibJS/Runtime/Temporal/Duration.cpp
+++ b/Userland/Libraries/LibJS/Runtime/Temporal/Duration.cpp
@@ -1071,10 +1071,12 @@ ThrowCompletionOr<DurationRecord> add_duration(GlobalObject& global_object, doub
         // a. Let diffNs be ! DifferenceInstant(relativeTo.[[Nanoseconds]], endNs, 1, "nanosecond", "halfExpand").
         auto* diff_ns = difference_instant(global_object, relative_to.nanoseconds(), *end_ns, 1, "nanosecond"sv, "halfExpand"sv);
 
-        // b. Let result be ! BalanceDuration(0, 0, 0, 0, 0, 0, diffNs, largestUnit).
+        // b. Assert: The following steps cannot fail due to overflow in the Number domain because abs(diffNs) ≤ 2 × nsMaxInstant.
+
+        // c. Let result be ! BalanceDuration(0, 0, 0, 0, 0, 0, diffNs, largestUnit).
         auto result = MUST(balance_duration(global_object, 0, 0, 0, 0, 0, 0, diff_ns->big_integer(), largest_unit));
 
-        // c. Return ? CreateDurationRecord(0, 0, 0, 0, result.[[Hours]], result.[[Minutes]], result.[[Seconds]], result.[[Milliseconds]], result.[[Microseconds]], result.[[Nanoseconds]]).
+        // d. Return ? CreateDurationRecord(0, 0, 0, 0, result.[[Hours]], result.[[Minutes]], result.[[Seconds]], result.[[Milliseconds]], result.[[Microseconds]], result.[[Nanoseconds]]).
         return create_duration_record(global_object, 0, 0, 0, 0, result.hours, result.minutes, result.seconds, result.milliseconds, result.microseconds, result.nanoseconds);
     }
 
diff --git a/Userland/Libraries/LibJS/Runtime/Temporal/Instant.cpp b/Userland/Libraries/LibJS/Runtime/Temporal/Instant.cpp
index 7eab596f06..4e3f5a2ba1 100644
--- a/Userland/Libraries/LibJS/Runtime/Temporal/Instant.cpp
+++ b/Userland/Libraries/LibJS/Runtime/Temporal/Instant.cpp
@@ -334,7 +334,7 @@ ThrowCompletionOr<Duration*> difference_temporal_instant(GlobalObject& global_ob
     // 12. Let roundedNs be ! DifferenceInstant(first.[[Nanoseconds]], second.[[Nanoseconds]], roundingIncrement, smallestUnit, roundingMode).
     auto* rounded_ns = difference_instant(global_object, first->nanoseconds(), second->nanoseconds(), rounding_increment, *smallest_unit, rounding_mode);
 
-    // 13. Assert: The following steps cannot fail due to overflow in the Number domain because abs(roundedNs) ≤ 1.728 × 10^22.
+    // 13. Assert: The following steps cannot fail due to overflow in the Number domain because abs(roundedNs) ≤ 2 × nsMaxInstant.
 
     // 14. Let result be ! BalanceDuration(0, 0, 0, 0, 0, 0, roundedNs, largestUnit).
     auto result = MUST(balance_duration(global_object, 0, 0, 0, 0, 0, 0, rounded_ns->big_integer(), *largest_unit));
diff --git a/Userland/Libraries/LibJS/Runtime/Temporal/ZonedDateTime.cpp b/Userland/Libraries/LibJS/Runtime/Temporal/ZonedDateTime.cpp
index 74037236c0..5069d2d9e5 100644
--- a/Userland/Libraries/LibJS/Runtime/Temporal/ZonedDateTime.cpp
+++ b/Userland/Libraries/LibJS/Runtime/Temporal/ZonedDateTime.cpp
@@ -589,7 +589,7 @@ ThrowCompletionOr<Duration*> difference_temporal_zoned_date_time(GlobalObject& g
         // a. Let differenceNs be ! DifferenceInstant(zonedDateTime.[[Nanoseconds]], other.[[Nanoseconds]], roundingIncrement, smallestUnit, roundingMode).
         auto* difference_ns = difference_instant(global_object, zoned_date_time.nanoseconds(), other->nanoseconds(), rounding_increment, *smallest_unit, rounding_mode);
 
-        // b. Assert: The following steps cannot fail due to overflow in the Number domain because abs(differenceNs) ≤ 1.728 × 10^22.
+        // b. Assert: The following steps cannot fail due to overflow in the Number domain because abs(differenceNs) ≤ 2 × nsMaxInstant.
 
         // c. Let balanceResult be ! BalanceDuration(0, 0, 0, 0, 0, 0, differenceNs, largestUnit).
         auto balance_result = MUST(balance_duration(global_object, 0, 0, 0, 0, 0, 0, difference_ns->big_integer(), *largest_unit));