diff --git a/tests/test_core/test_utils/test_string_utils.py b/tests/test_core/test_utils/test_string_utils.py
new file mode 100644
index 00000000000..7eea7732b58
--- /dev/null
+++ b/tests/test_core/test_utils/test_string_utils.py
@@ -0,0 +1,127 @@
+"""Unit tests for the pure string helpers in ``_plotly_utils.utils``.
+
+These functions back error messages that point at the offending part of a
+property path (split the path, find character positions, mark them, and
+suggest the closest valid key). They had no direct test coverage.
+"""
+
+from _plotly_utils.utils import (
+    _natural_sort_strings,
+    chomp_empty_strings,
+    display_string_positions,
+    find_closest_string,
+    levenshtein,
+    split_multichar,
+    split_string_positions,
+)
+
+SAMPLE = "a.string[0].with_separators"
+SEPARATORS = ".[]_"
+
+
+def test_split_multichar_splits_on_every_char():
+    assert split_multichar([SAMPLE], list(SEPARATORS)) == [
+        "a",
+        "string",
+        "0",
+        "",
+        "with",
+        "separators",
+    ]
+
+
+def test_split_multichar_no_chars_returns_input():
+    assert split_multichar([SAMPLE], []) == [SAMPLE]
+
+
+def test_split_string_positions():
+    parts = split_multichar([SAMPLE], list(SEPARATORS))
+    assert split_string_positions(parts) == [0, 2, 9, 11, 12, 17]
+
+
+def test_display_string_positions_single_index():
+    pos = split_string_positions(split_multichar([SAMPLE], list(SEPARATORS)))
+    assert display_string_positions(pos, 4) == "            ^"
+
+
+def test_display_string_positions_offset_length_char_no_trim():
+    pos = split_string_positions(split_multichar([SAMPLE], list(SEPARATORS)))
+    assert (
+        display_string_positions(pos, 4, offset=1, length=3, char="~", trim=False)
+        == "             ~~~      "
+    )
+
+
+def test_display_string_positions_all_indices():
+    pos = split_string_positions(split_multichar([SAMPLE], list(SEPARATORS)))
+    assert display_string_positions(pos) == "^ ^      ^ ^^    ^"
+
+
+def test_chomp_empty_strings_basic():
+    assert chomp_empty_strings(["hey", "", "why", "", "", "whoa", "", ""], "_") == [
+        "hey_",
+        "why__",
+        "whoa__",
+    ]
+
+
+def test_chomp_empty_strings_leading_empty():
+    assert chomp_empty_strings(["", "hi", "", "I'm", "bob", "", ""], "_") == [
+        "_",
+        "hi_",
+        "I'm",
+        "bob__",
+    ]
+
+
+def test_chomp_empty_strings_no_empties_unchanged():
+    strings = ["hi", "i'm", "a", "good", "string"]
+    assert chomp_empty_strings(strings, "_") == strings
+
+
+def test_chomp_empty_strings_special_cases():
+    assert chomp_empty_strings([], "_") == []
+    assert chomp_empty_strings([""], "_") == [""]
+    assert chomp_empty_strings(["", ""], "_") == ["_"]
+    assert chomp_empty_strings(["", "", "", ""], "_") == ["___"]
+
+
+def test_chomp_empty_strings_reverse():
+    # In reverse, an empty string attaches to the neighbor on its right.
+    assert chomp_empty_strings(["hey", "", "why"], "_", reverse=True) == [
+        "hey",
+        "_why",
+    ]
+
+
+def test_levenshtein():
+    assert levenshtein("kitten", "sitting") == 3
+    assert levenshtein("flaw", "lawn") == 2
+    assert levenshtein("abc", "abc") == 0
+    assert levenshtein("", "abc") == 3
+    # symmetric
+    assert levenshtein("abc", "") == levenshtein("", "abc")
+
+
+def test_find_closest_string():
+    assert find_closest_string("helo", ["hello", "world", "help"]) == "hello"
+
+
+def test_find_closest_string_ties_break_lexicographically():
+    # "ax" and "ay" are both distance 1 from "aa"; the lexicographically
+    # smaller one wins for a stable result.
+    assert find_closest_string("aa", ["ay", "ax"]) == "ax"
+
+
+def test_natural_sort_strings():
+    assert _natural_sort_strings(["a10", "a2", "a1", "a10b"]) == [
+        "a1",
+        "a2",
+        "a10",
+        "a10b",
+    ]
+    assert _natural_sort_strings(["a1", "a2", "a10"], reverse=True) == [
+        "a10",
+        "a2",
+        "a1",
+    ]