diff --git a/.github/workflows/ci-gpu.yml b/.github/workflows/ci-gpu.yml
index d2a2e95..3c3e193 100644
--- a/.github/workflows/ci-gpu.yml
+++ b/.github/workflows/ci-gpu.yml
@@ -43,9 +43,9 @@ jobs:
       - uses: actions/setup-python@v5
         with:
           python-version: "3.12"
-      - uses: hynek/setup-cached-uv@v2
+      - uses: astral-sh/setup-uv@v6
         with:
-          cache-dependency-path: pyproject.toml
+          enable-cache: true  # "auto" is `false` on non-GitHub runners
       - name: Install package
         run: uv pip install --system -e .[test,full] cupy-cuda12x --extra-index-url=https://pypi.nvidia.com --index-strategy=unsafe-best-match
       - name: List installed packages
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index d449dec..a4e24fe 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -75,7 +75,7 @@ jobs:
     steps:
       - uses: actions/checkout@v5
         with: { fetch-depth: 0, filter: "blob:none" }
-      - uses: actions/setup-python@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: '3.13'
       - uses: astral-sh/setup-uv@v7
@@ -90,7 +90,7 @@ jobs:
     steps:
       - uses: actions/checkout@v5
         with: { fetch-depth: 0, filter: "blob:none" }
-      - uses: actions/setup-python@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: '3.13'
       - uses: astral-sh/setup-uv@v7
@@ -110,7 +110,7 @@ jobs:
     steps:
       - uses: actions/checkout@v5
         with: { fetch-depth: 0, filter: "blob:none" }
-      - uses: actions/setup-python@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: ${{ matrix.python-version }}
       - uses: pre-commit/action@v3.0.1
diff --git a/.github/workflows/publish.yml b/.github/workflows/publish.yml
index 74c5322..d559a51 100644
--- a/.github/workflows/publish.yml
+++ b/.github/workflows/publish.yml
@@ -13,8 +13,8 @@ jobs:
     permissions:
       id-token: write # to authenticate as Trusted Publisher to pypi.org
     steps:
-      - uses: actions/checkout@v4
-      - uses: actions/setup-python@v5
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: "3.x"
           cache: pip
diff --git a/.readthedocs.yml b/.readthedocs.yml
index ca1d026..f32ea6c 100644
--- a/.readthedocs.yml
+++ b/.readthedocs.yml
@@ -2,7 +2,7 @@ version: 2
 build:
   os: ubuntu-24.04
   tools:
-    python: "3.12"
+    python: "3.13"
 python:
   install:
     - method: pip
diff --git a/docs/conf.py b/docs/conf.py
index 1c04e72..657de4e 100644
--- a/docs/conf.py
+++ b/docs/conf.py
@@ -9,16 +9,20 @@
 from typing import TYPE_CHECKING
 
 from docutils.nodes import Text
+from sphinx.util import logging
 
 
 if TYPE_CHECKING:
+    from typing import Final
+
     from docutils.nodes import TextElement, reference
     from sphinx.addnodes import pending_xref
     from sphinx.application import Sphinx
     from sphinx.environment import BuildEnvironment
 
 
-HERE = Path(__file__).parent
+HERE: Final = Path(__file__).parent
+LOGGER: Final = logging.getLogger("fast_array_utils")
 
 
 # -- General configuration ------------------------------------------------
@@ -134,7 +138,8 @@ def resolve_type_aliases(app: Sphinx, env: BuildEnvironment, node: pending_xref,
         ref = resolve_reference_any_inventory(env=env, honor_disabled_refs=False, node=node, contnode=contnode)
     if ref is None:
         msg = f"Could not resolve {typ} {target} (from {node['reftarget']})"
-        raise AssertionError(msg)
+        LOGGER.warning(msg, type="ref")
+        return ref
     if name:
         ref.children[:] = [Text(name)]
     return ref
diff --git a/pyproject.toml b/pyproject.toml
index edc5f27..ae26151 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -14,12 +14,11 @@ license = "MPL-2.0"
 authors = [
   { name = "Philipp A.", email = "flying-sheep@web.de" },
 ]
-requires-python = ">=3.11"
+requires-python = ">=3.12"
 classifiers = [
   "Development Status :: 5 - Production/Stable",
   "Intended Audience :: Developers",
   "Programming Language :: Python :: 3 :: Only",
-  "Programming Language :: Python :: 3.11",
   "Programming Language :: Python :: 3.12",
   "Programming Language :: Python :: 3.13",
   "Programming Language :: Python :: 3.14",
@@ -107,11 +106,11 @@ overrides.matrix.resolution.dependencies = [
 ]
 
 [[tool.hatch.envs.hatch-test.matrix]]
-python = [ "3.13", "3.11" ]
+python = [ "3.13", "3.12" ]
 extras = [ "full", "min" ]
 
 [[tool.hatch.envs.hatch-test.matrix]]
-python = [ "3.11" ]
+python = [ "3.12" ]
 extras = [ "full" ]
 resolution = [ "lowest" ]
 
diff --git a/src/fast_array_utils/conv/__init__.py b/src/fast_array_utils/conv/__init__.py
index 58afd56..6a34ca0 100644
--- a/src/fast_array_utils/conv/__init__.py
+++ b/src/fast_array_utils/conv/__init__.py
@@ -5,7 +5,6 @@
 
 from typing import TYPE_CHECKING, overload
 
-from ..typing import CpuArray, DiskArray, GpuArray  # noqa: TC001
 from ._to_dense import to_dense_
 
 
@@ -15,6 +14,7 @@
     from numpy.typing import NDArray
 
     from .. import types
+    from ..typing import CpuArray, DiskArray, GpuArray
 
 
 __all__ = ["to_dense"]
diff --git a/src/fast_array_utils/conv/_to_dense.py b/src/fast_array_utils/conv/_to_dense.py
index 249ec07..6b5de88 100644
--- a/src/fast_array_utils/conv/_to_dense.py
+++ b/src/fast_array_utils/conv/_to_dense.py
@@ -8,7 +8,6 @@
 import numpy as np
 
 from .. import types
-from ..typing import CpuArray, DiskArray, GpuArray  # noqa: TC001
 
 
 if TYPE_CHECKING:
@@ -16,6 +15,8 @@
 
     from numpy.typing import NDArray
 
+    from ..typing import CpuArray, DiskArray, GpuArray
+
 
 # fallback’s arg0 type has to include types of registered functions
 @singledispatch
diff --git a/src/fast_array_utils/stats/__init__.py b/src/fast_array_utils/stats/__init__.py
index 8010444..24712d8 100644
--- a/src/fast_array_utils/stats/__init__.py
+++ b/src/fast_array_utils/stats/__init__.py
@@ -10,7 +10,6 @@
 from typing import TYPE_CHECKING, cast, get_args, overload
 
 from .._validation import validate_axis
-from ..typing import CpuArray, DiskArray, GpuArray  # noqa: TC001
 from ._generic_ops import DtypeOps
 
 
@@ -22,6 +21,7 @@
     from optype.numpy import ToDType
 
     from .. import types
+    from ..typing import CpuArray, DiskArray, GpuArray
     from ._generic_ops import Ops
     from ._typing import NoDtypeOps, StatFunDtype, StatFunNoDtype
 
diff --git a/src/fast_array_utils/stats/_generic_ops.py b/src/fast_array_utils/stats/_generic_ops.py
index ef833ed..79df974 100644
--- a/src/fast_array_utils/stats/_generic_ops.py
+++ b/src/fast_array_utils/stats/_generic_ops.py
@@ -12,14 +12,14 @@
 
 
 if TYPE_CHECKING:
-    from typing import Any, Literal, TypeAlias
+    from typing import Any, Literal
 
     from numpy.typing import DTypeLike, NDArray
 
     from ..typing import CpuArray, DiskArray, GpuArray
     from ._typing import Ops
 
-    ComplexAxis: TypeAlias = tuple[Literal[0], Literal[1]] | tuple[Literal[0, 1]] | Literal[0, 1] | None
+    type ComplexAxis = tuple[Literal[0], Literal[1]] | tuple[Literal[0, 1]] | Literal[0, 1] | None
 
 
 def _run_numpy_op(
diff --git a/src/fast_array_utils/stats/_is_constant.py b/src/fast_array_utils/stats/_is_constant.py
index 4df5efd..1ac95d3 100644
--- a/src/fast_array_utils/stats/_is_constant.py
+++ b/src/fast_array_utils/stats/_is_constant.py
@@ -11,13 +11,10 @@
 
 
 if TYPE_CHECKING:
-    from collections.abc import Callable
-    from typing import Any, Literal, TypeVar
+    from typing import Any, Literal
 
     from numpy.typing import NDArray
 
-    C = TypeVar("C", bound=Callable[..., Any])
-
 
 @singledispatch
 def is_constant_(
diff --git a/src/fast_array_utils/stats/_power.py b/src/fast_array_utils/stats/_power.py
index 43dada2..8387836 100644
--- a/src/fast_array_utils/stats/_power.py
+++ b/src/fast_array_utils/stats/_power.py
@@ -10,20 +10,15 @@
 
 
 if TYPE_CHECKING:
-    from typing import TypeAlias, TypeVar
-
     from numpy.typing import DTypeLike
 
     from fast_array_utils.typing import CpuArray, GpuArray
 
     # All supported array types except for disk ones and CSDataset
-    Array: TypeAlias = CpuArray | GpuArray | types.DaskArray
-
-    _Arr = TypeVar("_Arr", bound=Array)
-    _Mat = TypeVar("_Mat", bound=types.CSBase | types.CupyCSMatrix)
+    type Array = CpuArray | GpuArray | types.DaskArray
 
 
-def power(x: _Arr, n: int, /, dtype: DTypeLike | None = None) -> _Arr:
+def power[Arr: Array](x: Arr, n: int, /, dtype: DTypeLike | None = None) -> Arr:
     """Take array or matrix to a power."""
     # This wrapper is necessary because TypeVars can’t be used in `singledispatch` functions
     return _power(x, n, dtype=dtype)  # type: ignore[return-value]
@@ -37,7 +32,7 @@ def _power(x: Array, n: int, /, dtype: DTypeLike | None = None) -> Array:
 
 
 @_power.register(types.CSBase | types.CupyCSMatrix)
-def _power_cs(x: _Mat, n: int, /, dtype: DTypeLike | None = None) -> _Mat:
+def _power_cs[Mat: types.CSBase | types.CupyCSMatrix](x: Mat, n: int, /, dtype: DTypeLike | None = None) -> Mat:
     new_data = power(x.data, n, dtype=dtype)
     return type(x)((new_data, x.indices, x.indptr), shape=x.shape, dtype=new_data.dtype)  # type: ignore[call-overload,return-value]
 
diff --git a/src/fast_array_utils/stats/_typing.py b/src/fast_array_utils/stats/_typing.py
index 2ddc4c0..be671dd 100644
--- a/src/fast_array_utils/stats/_typing.py
+++ b/src/fast_array_utils/stats/_typing.py
@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: MPL-2.0
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Generic, Literal, Protocol, TypedDict, TypeVar
+from typing import TYPE_CHECKING, Literal, Protocol, TypedDict
 
 import numpy as np
 
@@ -11,17 +11,17 @@
 
 
 if TYPE_CHECKING:
-    from typing import Any, TypeAlias
+    from typing import Any
 
     from numpy.typing import DTypeLike, NDArray
 
 
-Array: TypeAlias = CpuArray | GpuArray | DiskArray | types.CSDataset | types.DaskArray
+type Array = CpuArray | GpuArray | DiskArray | types.CSDataset | types.DaskArray
 
-DTypeIn: TypeAlias = np.float32 | np.float64 | np.int32 | np.bool_
-DTypeOut: TypeAlias = np.float32 | np.float64 | np.int64
+type DTypeIn = np.float32 | np.float64 | np.int32 | np.bool_
+type DTypeOut = np.float32 | np.float64 | np.int64
 
-NdAndAx: TypeAlias = tuple[Literal[1], None] | tuple[Literal[2], Literal[0, 1] | None]
+type NdAndAx = tuple[Literal[1], None] | tuple[Literal[2], Literal[0, 1] | None]
 
 
 class StatFunNoDtype(Protocol):
@@ -48,11 +48,8 @@ def __call__(
 
 NoDtypeOps = Literal["max", "min"]
 DtypeOps = Literal["sum"]
-Ops: TypeAlias = NoDtypeOps | DtypeOps
+type Ops = NoDtypeOps | DtypeOps
 
 
-_DT = TypeVar("_DT", bound="DTypeLike")
-
-
-class DTypeKw(TypedDict, Generic[_DT], total=False):
-    dtype: _DT
+class DTypeKw[DT: DTypeLike](TypedDict, total=False):
+    dtype: DT
diff --git a/src/fast_array_utils/stats/_utils.py b/src/fast_array_utils/stats/_utils.py
index 1ca4065..bc1f650 100644
--- a/src/fast_array_utils/stats/_utils.py
+++ b/src/fast_array_utils/stats/_utils.py
@@ -2,25 +2,24 @@
 from __future__ import annotations
 
 from functools import partial
-from typing import TYPE_CHECKING, Literal, TypeVar, cast, get_args
+from typing import TYPE_CHECKING, Literal, cast, get_args
 
 import numpy as np
 from numpy.exceptions import AxisError
 
 from .. import types
-from ..typing import GpuArray
 from ._typing import DtypeOps
 
 
 if TYPE_CHECKING:
-    from typing import Any, Literal, TypeAlias
+    from typing import Any, Literal
 
     from numpy.typing import DTypeLike, NDArray
 
-    from ..typing import CpuArray
+    from ..typing import CpuArray, GpuArray
     from ._typing import DTypeKw, Ops
 
-    ComplexAxis: TypeAlias = tuple[Literal[0], Literal[1]] | tuple[Literal[0, 1]] | Literal[0, 1] | None
+    type ComplexAxis = tuple[Literal[0], Literal[1]] | tuple[Literal[0, 1]] | Literal[0, 1] | None
 
 
 __all__ = ["_dask_inner"]
@@ -71,7 +70,7 @@ def _dask_block(
     from . import max, min, sum
 
     if computing_meta:  # dask.blockwise doesn’t allow to pass `meta` in, and reductions below don’t handle a 0d matrix
-        return (types.CupyArray if isinstance(a, GpuArray) else np.ndarray)((), dtype or a.dtype)
+        return (types.CupyArray if isinstance(a, types.CupyArray | types.CupyCSMatrix) else np.ndarray)((), dtype or a.dtype)
 
     fns = {fn.__name__: fn for fn in (min, max, sum)}
 
@@ -114,8 +113,5 @@ def _get_shape(a: NDArray[Any] | np.number[Any] | types.CupyArray, *, axis: Lite
             raise AssertionError(msg)
 
 
-DT = TypeVar("DT", bound="DTypeLike")
-
-
-def _dtype_kw(dtype: DT | None, op: Ops) -> DTypeKw[DT]:
+def _dtype_kw[DT: DTypeLike](dtype: DT | None, op: Ops) -> DTypeKw[DT]:
     return {"dtype": dtype} if dtype is not None and op in get_args(DtypeOps) else {}
diff --git a/src/fast_array_utils/types.py b/src/fast_array_utils/types.py
index 6269c69..41e6c6f 100644
--- a/src/fast_array_utils/types.py
+++ b/src/fast_array_utils/types.py
@@ -4,7 +4,7 @@
 from __future__ import annotations
 
 from importlib.util import find_spec
-from typing import TYPE_CHECKING, TypeVar
+from typing import TYPE_CHECKING
 
 
 __all__ = [
@@ -34,8 +34,6 @@
     "spmatrix",
 ]
 
-T_co = TypeVar("T_co", covariant=True)
-
 
 # scipy sparse
 if TYPE_CHECKING:
diff --git a/src/fast_array_utils/typing.py b/src/fast_array_utils/typing.py
index af7238b..aebcdbb 100644
--- a/src/fast_array_utils/typing.py
+++ b/src/fast_array_utils/typing.py
@@ -16,13 +16,14 @@
 
 __all__ = ["CpuArray", "DiskArray", "GpuArray"]
 
+# change to `type` syntax once this is released: https://github.com/sphinx-doc/sphinx/pull/13508
 
-CpuArray: TypeAlias = NDArray[Any] | types.CSBase
+CpuArray: TypeAlias = NDArray[Any] | types.CSBase  # noqa: UP040
 """Arrays and matrices stored in CPU memory."""
 
-GpuArray: TypeAlias = types.CupyArray | types.CupyCSMatrix
+GpuArray: TypeAlias = types.CupyArray | types.CupyCSMatrix  # noqa: UP040
 """Arrays and matrices stored in GPU memory."""
 
 # TODO(flying-sheep): types.CSDataset  # noqa: TD003
-DiskArray: TypeAlias = types.H5Dataset | types.ZarrArray
+DiskArray: TypeAlias = types.H5Dataset | types.ZarrArray  # noqa: UP040
 """Arrays and matrices stored on disk."""
diff --git a/src/testing/fast_array_utils/_array_type.py b/src/testing/fast_array_utils/_array_type.py
index 48c0280..b5649a3 100644
--- a/src/testing/fast_array_utils/_array_type.py
+++ b/src/testing/fast_array_utils/_array_type.py
@@ -4,6 +4,7 @@
 from __future__ import annotations
 
 import enum
+import sys
 from dataclasses import KW_ONLY, dataclass, field
 from functools import cached_property, partial
 from importlib.metadata import version
@@ -17,23 +18,18 @@
 
 
 if TYPE_CHECKING:
-    from typing import Any, Literal, Protocol, TypeAlias
+    from typing import Any, Literal, Protocol
 
     import h5py
     from numpy.typing import ArrayLike, DTypeLike, NDArray
 
     from fast_array_utils.typing import CpuArray, DiskArray, GpuArray
 
-    InnerArray = CpuArray | GpuArray | DiskArray
-    Array: TypeAlias = InnerArray | types.DaskArray | types.CSDataset
-    ExtendedArray = Array | types.COOBase | types.CupyCOOMatrix
+    type InnerArray = CpuArray | GpuArray | DiskArray
+    type Array = InnerArray | types.DaskArray | types.CSDataset
+    type ExtendedArray = Array | types.COOBase | types.CupyCOOMatrix
 
-    Arr = TypeVar("Arr", bound=ExtendedArray, default=Array)
-    Arr_co = TypeVar("Arr_co", bound=ExtendedArray, covariant=True)
-
-    Inner = TypeVar("Inner", bound="ArrayType[InnerArray, None] | None", default=Any)
-
-    class ToArray(Protocol, Generic[Arr_co]):
+    class ToArray[Arr_co: ExtendedArray](Protocol):
         """Convert to a supported array."""
 
         def __call__(self, data: ArrayLike | Array, /, *, dtype: DTypeLike | None = None) -> Arr_co: ...
@@ -41,14 +37,11 @@ def __call__(self, data: ArrayLike | Array, /, *, dtype: DTypeLike | None = None
     class MkArray(Protocol):
         def __call__(self, shape: tuple[int, int], /, *, dtype: DTypeLike | None = None) -> Array: ...
 
-    _DTypeLikeFloat32 = np.dtype[np.float32] | type[np.float32]
-    _DTypeLikeFloat64 = np.dtype[np.float64] | type[np.float64]
-    _DTypeLikeInt32 = np.dtype[np.int32] | type[np.int32]
-    _DTypeLikeIn64 = np.dtype[np.int64] | type[np.int64]
-    _DTypeLikeNum = _DTypeLikeFloat32 | _DTypeLikeFloat64 | _DTypeLikeInt32 | _DTypeLikeIn64
-else:
-    Arr = TypeVar("Arr")
-    Inner = TypeVar("Inner")
+    type _DTypeLikeFloat32 = np.dtype[np.float32] | type[np.float32]
+    type _DTypeLikeFloat64 = np.dtype[np.float64] | type[np.float64]
+    type _DTypeLikeInt32 = np.dtype[np.int32] | type[np.int32]
+    type _DTypeLikeIn64 = np.dtype[np.int64] | type[np.int64]
+    type _DTypeLikeNum = _DTypeLikeFloat32 | _DTypeLikeFloat64 | _DTypeLikeInt32 | _DTypeLikeIn64
 
 
 __all__ = ["ArrayType", "ConversionContext", "ToArray"]
@@ -81,8 +74,17 @@ class ConversionContext:
     hdf5_file: h5py.File  # TODO(flying-sheep): ReadOnly <https://peps.python.org/pep-0767/>
 
 
+if TYPE_CHECKING or sys.version_info >= (3, 13):
+    # TODO(flying-sheep): move vars into type parameter syntax  # noqa: TD003
+    Arr = TypeVar("Arr", bound="ExtendedArray", default="Array")
+    Inner = TypeVar("Inner", bound="ArrayType[InnerArray, None] | None", default="Any")
+else:
+    Arr = TypeVar("Arr")
+    Inner = TypeVar("Inner")
+
+
 @dataclass(frozen=True)
-class ArrayType(Generic[Arr, Inner]):
+class ArrayType(Generic[Arr, Inner]):  # noqa: UP046
     """Supported array type with methods for conversion and random generation.
 
     Examples
diff --git a/tests/test_to_dense.py b/tests/test_to_dense.py
index a7b5692..1d75969 100644
--- a/tests/test_to_dense.py
+++ b/tests/test_to_dense.py
@@ -14,13 +14,13 @@
 
 if TYPE_CHECKING:
     from collections.abc import Iterable
-    from typing import Literal, TypeAlias
+    from typing import Literal
 
     from fast_array_utils.typing import CpuArray, DiskArray, GpuArray
     from testing.fast_array_utils import ArrayType
 
-    Array: TypeAlias = CpuArray | GpuArray | DiskArray | types.CSDataset | types.DaskArray
-    ExtendedArray: TypeAlias = Array | types.COOBase | types.CupyCOOMatrix
+    type Array = CpuArray | GpuArray | DiskArray | types.CSDataset | types.DaskArray
+    type ExtendedArray = Array | types.COOBase | types.CupyCOOMatrix
 
 
 WARNS_NUMBA = pytest.warns(RuntimeWarning, match="numba is not installed; falling back to slow conversion")
diff --git a/typings/dask/array/core.pyi b/typings/dask/array/core.pyi
index dba0e6c..7f65dd2 100644
--- a/typings/dask/array/core.pyi
+++ b/typings/dask/array/core.pyi
@@ -1,19 +1,18 @@
 # SPDX-License-Identifier: MPL-2.0
 # pyright: reportIncompatibleMethodOverride=false
 from collections.abc import Callable, Sequence
-from typing import Any, Literal, Never, TypeAlias
+from typing import Any, Literal, Never, override
 
 import cupy
 import cupyx.scipy.sparse
 import numpy as np
 import scipy.sparse
 from numpy.typing import DTypeLike, NDArray
-from typing_extensions import override
 
 from ..utils import SerializableLock
 
-_Chunks: TypeAlias = tuple[int, ...] | tuple[tuple[int, ...], ...]
-_Array: TypeAlias = (
+type _Chunks = tuple[int, ...] | tuple[tuple[int, ...], ...]
+type _Array = (
     NDArray[Any]
     | scipy.sparse.csr_array
     | scipy.sparse.csc_array
diff --git a/typings/dask/array/overlap.pyi b/typings/dask/array/overlap.pyi
index 73d00ad..ce1a65a 100644
--- a/typings/dask/array/overlap.pyi
+++ b/typings/dask/array/overlap.pyi
@@ -1,12 +1,12 @@
 # SPDX-License-Identifier: MPL-2.0
 from collections.abc import Callable
-from typing import Literal, TypeAlias
+from typing import Literal
 
 from .core import Array, _Array
 
-_Depth: TypeAlias = int | tuple[int, ...] | dict[int, _Depth]
-_Boundary: TypeAlias = Literal["reflect", "periodic", "nearest", "none"] | int
-_Boundaries: TypeAlias = _Boundary | tuple[_Boundary, ...] | dict[int, _Boundary]
+type _Depth = int | tuple[int, ...] | dict[int, _Depth]
+type _Boundary = Literal["reflect", "periodic", "nearest", "none"] | int
+type _Boundaries = _Boundary | tuple[_Boundary, ...] | dict[int, _Boundary]
 
 def map_overlap(
     func: Callable[[_Array], _Array],
diff --git a/typings/dask/array/wrap.pyi b/typings/dask/array/wrap.pyi
index 497950d..1559667 100644
--- a/typings/dask/array/wrap.pyi
+++ b/typings/dask/array/wrap.pyi
@@ -1,13 +1,13 @@
 # SPDX-License-Identifier: MPL-2.0
 
-from typing import Any, Literal, TypeAlias
+from typing import Any, Literal
 
 import numpy as np
 from numpy.typing import ArrayLike, DTypeLike
 
 from .core import Array, _Chunks
 
-_Order: TypeAlias = Literal["C", "F", "A", "K"]
+type _Order = Literal["C", "F", "A", "K"]
 
 def full(
     a: ArrayLike | None = None,
diff --git a/typings/numba/__init__.pyi b/typings/numba/__init__.pyi
index 016adff..c3a1395 100644
--- a/typings/numba/__init__.pyi
+++ b/typings/numba/__init__.pyi
@@ -1,18 +1,17 @@
 # SPDX-License-Identifier: MPL-2.0
 from collections.abc import Callable, Iterable
-from typing import Literal, SupportsIndex, TypeAlias, TypeVar, overload
+from typing import Literal, SupportsIndex, overload
 
 from .core.types import *
 
-_F = TypeVar("_F", bound=Callable[..., object])
-
-_Signature: TypeAlias = str | Type | tuple[_Signature, ...]
+type __Signature = str | Type
+type _Signature = str | Type | tuple[__Signature, ...]
 
 # https://numba.readthedocs.io/en/stable/reference/jit-compilation.html#numba.jit
 @overload
-def njit(f: _F) -> _F: ...
+def njit[F: Callable[..., object]](f: F) -> F: ...
 @overload
-def njit(
+def njit[F: Callable[..., object]](
     signature: _Signature | list[_Signature] | None = None,
     *,
     nopython: bool = True,
@@ -24,7 +23,7 @@ def njit(
     fastmath: bool = False,
     locals: dict[str, object] = {},
     boundscheck: bool = False,
-) -> Callable[[_F], _F]: ...
+) -> Callable[[F], F]: ...
 @overload
 def prange(stop: SupportsIndex, /) -> Iterable[int]: ...
 @overload
diff --git a/typings/numba/core/datamodel/new_models.pyi b/typings/numba/core/datamodel/new_models.pyi
index dd5c0a0..3b79798 100644
--- a/typings/numba/core/datamodel/new_models.pyi
+++ b/typings/numba/core/datamodel/new_models.pyi
@@ -1,14 +1,11 @@
 # SPDX-License-Identifier: MPL-2.0
 from collections.abc import Iterable
-from typing import Generic, TypeVar
 
 from ..types import Type
 from .manager import DataModelManager
 
-_T = TypeVar("_T", bound=Type)
+class DataModel[T: Type]: ...
+class CompositeModel[T: Type](DataModel[T]): ...
 
-class DataModel(Generic[_T]): ...
-class CompositeModel(DataModel[_T], Generic[_T]): ...
-
-class StructModel(CompositeModel[_T], Generic[_T]):
-    def __init__(self, dmm: DataModelManager, fe_type: _T, members: Iterable[tuple[str, Type]]) -> None: ...
+class StructModel[T: Type](CompositeModel[T]):
+    def __init__(self, dmm: DataModelManager, fe_type: T, members: Iterable[tuple[str, Type]]) -> None: ...
diff --git a/typings/numba/core/datamodel/registry.pyi b/typings/numba/core/datamodel/registry.pyi
index 3ff8ac3..d15ce3b 100644
--- a/typings/numba/core/datamodel/registry.pyi
+++ b/typings/numba/core/datamodel/registry.pyi
@@ -1,9 +1,6 @@
 # SPDX-License-Identifier: MPL-2.0
 from collections.abc import Callable
-from typing import TypeVar
 
 from ..types import Type
 
-_F = TypeVar("_F", bound=Callable[..., object])
-
-def register_default(typecls: type[Type]) -> Callable[[_F], _F]: ...
+def register_default[F: Callable[..., object]](typecls: type[Type]) -> Callable[[F], F]: ...
diff --git a/typings/numba/core/extending.pyi b/typings/numba/core/extending.pyi
index bc9a639..5f6119f 100644
--- a/typings/numba/core/extending.pyi
+++ b/typings/numba/core/extending.pyi
@@ -2,7 +2,7 @@
 import typing
 from collections.abc import Callable
 from dataclasses import dataclass
-from typing import Concatenate, ParamSpec, Protocol, TypeVar
+from typing import Concatenate, Protocol
 
 from llvmlite.ir import IRBuilder, Value
 from numba.core.base import BaseContext
@@ -28,11 +28,6 @@ __all__ = [
     "unbox",
 ]
 
-_F = TypeVar("_F", bound=Callable[..., object])
-_P = ParamSpec("_P")
-_R = TypeVar("_R")
-_T = TypeVar("_T", bound=Type)
-
 TypingContext = object
 
 class _Context(Protocol):
@@ -55,33 +50,33 @@ class NativeValue:
     is_error: Value = ...
     cleanup: Value | None = None
 
-def box(
-    typeclass: type[_T],
+def box[T: Type, R](
+    typeclass: type[T],
 ) -> Callable[
-    [Callable[[_T, NativeValue, BoxContext], _R]],
-    Callable[[_T, NativeValue, BoxContext], _R],
+    [Callable[[T, NativeValue, BoxContext], R]],
+    Callable[[T, NativeValue, BoxContext], R],
 ]: ...
-def unbox(
-    typeclass: type[_T],
+def unbox[T: Type](
+    typeclass: type[T],
 ) -> Callable[
-    [Callable[[_T, Value, UnboxContext], NativeValue]],
-    Callable[[_T, Value, UnboxContext], NativeValue],
+    [Callable[[T, Value, UnboxContext], NativeValue]],
+    Callable[[T, Value, UnboxContext], NativeValue],
 ]: ...
 @typing.overload
-def intrinsic(
-    func: Callable[Concatenate[TypingContext, _P], tuple[Signature, Callable[..., NativeValue]]],
+def intrinsic[**P](
+    func: Callable[Concatenate[TypingContext, P], tuple[Signature, Callable[..., NativeValue]]],
     /,
-) -> Callable[_P, object]: ...
+) -> Callable[P, object]: ...
 @typing.overload
-def intrinsic(
+def intrinsic[**P](
     *,
     prefer_literal: bool = False,
     **kwargs: object,
 ) -> Callable[
-    [Callable[Concatenate[TypingContext, _P], tuple[Signature, Callable[..., NativeValue]]]],
-    Callable[_P, object],
+    [Callable[Concatenate[TypingContext, P], tuple[Signature, Callable[..., NativeValue]]]],
+    Callable[P, object],
 ]: ...
 def make_attribute_wrapper(typeclass: type[Type], struct_attr: str, python_attr: str) -> None: ...
-def overload(f: Callable[..., object]) -> Callable[[_F], _F]: ...
-def overload_method(typecls: type[Type], name: str) -> Callable[[_F], _F]: ...
-def overload_attribute(typecls: type[Type], name: str) -> Callable[[_F], _F]: ...
+def overload[F: Callable[..., object]](f: Callable[..., object]) -> Callable[[F], F]: ...
+def overload_method[F: Callable[..., object]](typecls: type[Type], name: str) -> Callable[[F], F]: ...
+def overload_attribute[F: Callable[..., object]](typecls: type[Type], name: str) -> Callable[[F], F]: ...
diff --git a/typings/numba/core/types.pyi b/typings/numba/core/types.pyi
index 2823b70..0660356 100644
--- a/typings/numba/core/types.pyi
+++ b/typings/numba/core/types.pyi
@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: MPL-2.0
 # See <https://numba.readthedocs.io/en/stable/reference/types.html#numba-types>
 
-from typing import Literal, Self, TypeAlias
+from typing import Literal, Self
 
 from numba.core.typing.templates import Signature
 
@@ -48,7 +48,7 @@ class ArrayCompatible(Type):
     dtype: Type
 
 class Buffer(IterableType, ArrayCompatible):
-    Layout: TypeAlias = Literal["C", "F", "CS", "FS", "A"]
+    type Layout = Literal["C", "F", "CS", "FS", "A"]
     LAYOUTS: frozenset[Layout]
     def copy(
         self,