diff --git a/src/numpy-stubs/@test/runtime/legacy/literal.py b/src/numpy-stubs/@test/runtime/legacy/literal.py
index 5369d404..a85f1e04 100644
--- a/src/numpy-stubs/@test/runtime/legacy/literal.py
+++ b/src/numpy-stubs/@test/runtime/legacy/literal.py
@@ -18,7 +18,7 @@
 CF = frozenset({None, "C", "F"})
 
 order_list: list[tuple[frozenset[str | None], Callable[..., Any]]] = [
-    (KACF, partial(np.ndarray, 1)),
+    (KACF, partial(np.ndarray.__call__, 1)),
     (KACF, AR.tobytes),
     (KACF, partial(AR.astype, int)),
     (KACF, AR.copy),
diff --git a/src/numpy-stubs/__init__.pyi b/src/numpy-stubs/__init__.pyi
index 1ac5e4d5..56f64c6d 100644
--- a/src/numpy-stubs/__init__.pyi
+++ b/src/numpy-stubs/__init__.pyi
@@ -619,7 +619,7 @@ _RealT_co = TypeVar("_RealT_co", covariant=True)
 _ImagT_co = TypeVar("_ImagT_co", covariant=True)
 
 _DTypeT = TypeVar("_DTypeT", bound=dtype)
-_DTypeT_co = TypeVar("_DTypeT_co", bound=dtype, covariant=True)
+_DTypeT_co = TypeVar("_DTypeT_co", bound=dtype, default=dtype, covariant=True)
 _FlexDTypeT = TypeVar("_FlexDTypeT", bound=dtype[flexible])
 
 _ArrayT = TypeVar("_ArrayT", bound=_nt.Array)
@@ -628,6 +628,7 @@ _NumericArrayT = TypeVar("_NumericArrayT", bound=_nt.Array[number | timedelta64
 
 _ShapeT = TypeVar("_ShapeT", bound=_nt.Shape)
 _ShapeT_co = TypeVar("_ShapeT_co", bound=_nt.Shape, covariant=True)
+_ShapeT0_co = TypeVar("_ShapeT0_co", bound=_nt.Shape, default=_nt.Shape, covariant=True)
 _Shape1NDT = TypeVar("_Shape1NDT", bound=_nt.Shape1N)
 
 _ScalarT = TypeVar("_ScalarT", bound=generic)
@@ -998,9 +999,11 @@ _HasTypeWithReal: TypeAlias = _HasType[_HasReal[_T]]
 _HasTypeWithImag: TypeAlias = _HasType[_HasImag[_T]]
 
 @type_check_only
-class _HasDType(Protocol[_T_co]):
+class _HasDType(Protocol[_T_co, _ShapeT0_co]):
     @property
     def dtype(self, /) -> _T_co: ...
+    @property
+    def shape(self, /) -> _ShapeT0_co: ...
 
 _HasDTypeWithItem: TypeAlias = _HasDType[_HasTypeWithItem[_T]]
 _HasDTypeWithReal: TypeAlias = _HasDType[_HasTypeWithReal[_T]]
@@ -2117,16 +2120,16 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     def __ge__(self, other: _ArrayLikeObject_co, /) -> _nt.Array[bool_]: ...
 
     #
-    def __abs__(self: _HasDType[_HasType[SupportsAbs[_ScalarT]]], /) -> _nt.Array[_ScalarT, _ShapeT_co]: ...
+    def __abs__(self: _HasDType[_HasType[SupportsAbs[_ScalarT]], _ShapeT], /) -> _nt.Array[_ScalarT, _ShapeT]: ...
     def __neg__(self: _NumericArrayT, /) -> _NumericArrayT: ...  # noqa: PYI019
     def __pos__(self: _NumericArrayT, /) -> _NumericArrayT: ...  # noqa: PYI019
     def __invert__(self: _IntegralArrayT, /) -> _IntegralArrayT: ...  # noqa: PYI019
 
     #
     @overload
-    def __add__(self: _nt.Array[_ScalarT], x: _nt.Casts[_ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __add__(self: _nt.Array[_ScalarT], x: _nt.Casts[_ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
-    def __add__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __add__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
     def __add__(self: _nt.CastsWithBuiltin[_T, _ScalarT], x: _nt.SequenceND[_T], /) -> _nt.Array[_ScalarT]: ...
     @overload
@@ -2138,11 +2141,11 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __add__(self: _nt.Array[datetime64], x: _nt.CoTimeDelta_nd, /) -> _nt.Array[datetime64]: ...
     @overload
-    def __add__(self: _nt.Array[_nt.co_timedelta], x: _nt.ToDateTime_nd, /) -> _nt.Array[datetime64]: ...
+    def __add__(self: _nt.Array[_nt.co_timedelta, Any], x: _nt.ToDateTime_nd, /) -> _nt.Array[datetime64]: ...
     @overload
-    def __add__(self: _nt.Array[object_], x: object, /) -> _nt.Array[object_]: ...  # type: ignore[overload-cannot-match]
+    def __add__(self: _nt.Array[object_, Any], x: object, /) -> _nt.Array[object_]: ...
     @overload
-    def __add__(self: _nt.Array[str_], x: _nt.ToString_nd[_T], /) -> _nt.StringArrayND[_T]: ...
+    def __add__(self: _nt.Array[str_, Any], x: _nt.ToString_nd[_T], /) -> _nt.StringArrayND[_T]: ...
     @overload
     def __add__(self: _nt.StringArrayND[_T], x: _nt.ToString_nd[_T] | _nt.ToStr_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
@@ -2150,9 +2153,9 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
 
     #
     @overload
-    def __radd__(self: _nt.Array[_ScalarT], x: _nt.Casts[_ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __radd__(self: _nt.Array[_ScalarT], x: _nt.Casts[_ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
-    def __radd__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __radd__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
     def __radd__(self: _nt.CastsWithBuiltin[_T, _ScalarT], x: _nt.SequenceND[_T], /) -> _nt.Array[_ScalarT]: ...
     @overload
@@ -2164,11 +2167,11 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __radd__(self: _nt.Array[datetime64], x: _nt.CoTimeDelta_nd, /) -> _nt.Array[datetime64]: ...
     @overload
-    def __radd__(self: _nt.Array[_nt.co_timedelta], x: _nt.ToDateTime_nd, /) -> _nt.Array[datetime64]: ...
+    def __radd__(self: _nt.Array[_nt.co_timedelta, Any], x: _nt.ToDateTime_nd, /) -> _nt.Array[datetime64]: ...
     @overload
-    def __radd__(self: _nt.Array[object_], x: object, /) -> _nt.Array[object_]: ...  # type: ignore[overload-cannot-match]
+    def __radd__(self: _nt.Array[object_, Any], x: object, /) -> _nt.Array[object_]: ...
     @overload
-    def __radd__(self: _nt.Array[str_], x: _nt.ToString_nd[_T], /) -> _nt.StringArrayND[_T]: ...
+    def __radd__(self: _nt.Array[str_, Any], x: _nt.ToString_nd[_T], /) -> _nt.StringArrayND[_T]: ...
     @overload
     def __radd__(self: _nt.StringArrayND[_T], x: _nt.ToString_nd[_T] | _nt.ToStr_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
@@ -2268,7 +2271,7 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __mul__(self: _nt.Array[_CoNumberT], x: _nt.Casts[_CoNumberT], /) -> _nt.Array[_CoNumberT]: ...
     @overload
-    def __mul__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __mul__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
     def __mul__(self: _nt.CastsWithBuiltin[_T, _ScalarT], x: _nt.SequenceND[_T], /) -> _nt.Array[_ScalarT]: ...
     @overload
@@ -2278,11 +2281,11 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __mul__(self: _nt.CastsWithComplex[_ScalarT], x: _PyComplexND, /) -> _nt.Array[_ScalarT]: ...
     @overload
-    def __mul__(self: _nt.Array[timedelta64], x: _nt.ToFloating_nd, /) -> _nt.Array[timedelta64]: ...
+    def __mul__(self: _nt.Array[timedelta64, Any], x: _nt.ToFloating_nd, /) -> _nt.Array[timedelta64]: ...
     @overload
-    def __mul__(self: _nt.Array[object_], x: object, /) -> _nt.Array[object_]: ...  # type: ignore[overload-cannot-match]
+    def __mul__(self: _nt.Array[object_, Any], x: object, /) -> _nt.Array[object_]: ...
     @overload
-    def __mul__(self: _nt.Array[integer], x: _nt.ToString_nd, /) -> _nt.StringArrayND[_T]: ...
+    def __mul__(self: _nt.Array[integer, Any], x: _nt.ToString_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
     def __mul__(self: _nt.StringArrayND[_T], x: _nt.ToInteger_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
@@ -2292,7 +2295,7 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __rmul__(self: _nt.Array[_CoNumberT], x: _nt.Casts[_CoNumberT], /) -> _nt.Array[_CoNumberT]: ...
     @overload
-    def __rmul__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...  # type: ignore[overload-overlap]
+    def __rmul__(self: _nt.Array[_SelfScalarT], x: _nt.CastsWith[_SelfScalarT, _ScalarT], /) -> _nt.Array[_ScalarT]: ...
     @overload
     def __rmul__(self: _nt.CastsWithBuiltin[_T, _ScalarT], x: _nt.SequenceND[_T], /) -> _nt.Array[_ScalarT]: ...
     @overload
@@ -2302,11 +2305,11 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
     @overload
     def __rmul__(self: _nt.CastsWithComplex[_ScalarT], x: _PyComplexND, /) -> _nt.Array[_ScalarT]: ...
     @overload
-    def __rmul__(self: _nt.Array[timedelta64], x: _nt.ToFloating_nd, /) -> _nt.Array[timedelta64]: ...
+    def __rmul__(self: _nt.Array[timedelta64, Any], x: _nt.ToFloating_nd, /) -> _nt.Array[timedelta64]: ...
     @overload
-    def __rmul__(self: _nt.Array[object_], x: object, /) -> _nt.Array[object_]: ...  # type: ignore[overload-cannot-match]
+    def __rmul__(self: _nt.Array[object_, Any], x: object, /) -> _nt.Array[object_]: ...
     @overload
-    def __rmul__(self: _nt.Array[integer], x: _nt.ToString_nd, /) -> _nt.StringArrayND[_T]: ...
+    def __rmul__(self: _nt.Array[integer, Any], x: _nt.ToString_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
     def __rmul__(self: _nt.StringArrayND[_T], x: _nt.ToInteger_nd, /) -> _nt.StringArrayND[_T]: ...
     @overload
@@ -2666,47 +2669,55 @@ class ndarray(_ArrayOrScalarCommon, Generic[_ShapeT_co, _DTypeT_co]):
 
     #
     @overload
-    def __divmod__(self: _nt.Array[bool_], x: _nt.ToBool_nd, /) -> _2Tuple[_nt.Array[int8]]: ...
+    def __divmod__(self: _nt.Array[bool_, _nt.Shape], x: _nt.ToBool_nd, /) -> _2Tuple[_nt.Array[int8]]: ...
     @overload
     def __divmod__(
-        self: _nt.Array[_RealNumberT], x: _nt.Casts[_RealNumberT] | _nt.ToBool_nd, /
+        self: _nt.Array[_RealNumberT, _nt.Shape], x: _nt.Casts[_RealNumberT, _nt.Shape] | _nt.ToBool_nd, /
     ) -> _2Tuple[_nt.Array[_RealNumberT]]: ...
     @overload
     def __divmod__(
-        self: _nt.Array[_CoFloatingT], x: _nt.CastsWith[_CoFloatingT, _RealScalarT], /
+        self: _nt.Array[_CoFloatingT, _nt.Shape], x: _nt.CastsWith[_CoFloatingT, _RealScalarT, _nt.Shape], /
     ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
-    def __divmod__(self: _nt.CastsWithInt[_RealScalarT], x: _PyIntND, /) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
+    def __divmod__(
+        self: _nt.CastsWithInt[_RealScalarT, _nt.Shape], x: _PyIntND, /
+    ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
-    def __divmod__(self: _nt.CastsWithFloat[_RealScalarT], x: _PyFloatND, /) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
+    def __divmod__(
+        self: _nt.CastsWithFloat[_RealScalarT, _nt.Shape], x: _PyFloatND, /
+    ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
     def __divmod__(
-        self: _nt.Array[timedelta64], x: _nt.ToTimeDelta_nd, /
+        self: _nt.Array[timedelta64, _nt.Shape], x: _nt.ToTimeDelta_nd, /
     ) -> tuple[_nt.Array[int64], _nt.Array[timedelta64]]: ...
     @overload
-    def __divmod__(self: _nt.Array[object_], x: object, /) -> _2Tuple[_nt.Array[object_]]: ...
+    def __divmod__(self: _nt.Array[object_, _nt.Shape], x: object, /) -> _2Tuple[_nt.Array[object_]]: ...
 
     #
     @overload
-    def __rdivmod__(self: _nt.Array[bool_], x: _nt.ToBool_nd, /) -> _2Tuple[_nt.Array[int8]]: ...
+    def __rdivmod__(self: _nt.Array[bool_, _nt.Shape], x: _nt.ToBool_nd, /) -> _2Tuple[_nt.Array[int8]]: ...
     @overload
     def __rdivmod__(
-        self: _nt.Array[_RealNumberT], x: _nt.Casts[_RealNumberT] | _nt.ToBool_nd, /
+        self: _nt.Array[_RealNumberT, _nt.Shape], x: _nt.Casts[_RealNumberT, _nt.Shape] | _nt.ToBool_nd, /
     ) -> _2Tuple[_nt.Array[_RealNumberT]]: ...
     @overload
     def __rdivmod__(
-        self: _nt.Array[_CoFloatingT], x: _nt.CastsWith[_CoFloatingT, _RealScalarT], /
+        self: _nt.Array[_CoFloatingT, _nt.Shape], x: _nt.CastsWith[_CoFloatingT, _RealScalarT, _nt.Shape], /
     ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
-    def __rdivmod__(self: _nt.CastsWithInt[_RealScalarT], x: _PyIntND, /) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
+    def __rdivmod__(
+        self: _nt.CastsWithInt[_RealScalarT, _nt.Shape], x: _PyIntND, /
+    ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
-    def __rdivmod__(self: _nt.CastsWithFloat[_RealScalarT], x: _PyFloatND, /) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
+    def __rdivmod__(
+        self: _nt.CastsWithFloat[_RealScalarT, _nt.Shape], x: _PyFloatND, /
+    ) -> _2Tuple[_nt.Array[_RealScalarT]]: ...
     @overload
     def __rdivmod__(
-        self: _nt.Array[timedelta64], x: _nt.ToTimeDelta_nd, /
+        self: _nt.Array[timedelta64, _nt.Shape], x: _nt.ToTimeDelta_nd, /
     ) -> tuple[_nt.Array[int64], _nt.Array[timedelta64]]: ...
     @overload
-    def __rdivmod__(self: _nt.Array[object_], x: object, /) -> _2Tuple[_nt.Array[object_]]: ...
+    def __rdivmod__(self: _nt.Array[object_, _nt.Shape], x: object, /) -> _2Tuple[_nt.Array[object_]]: ...
 
     #
     @overload