diff --git a/lamindb/__init__.py b/lamindb/__init__.py
index 3ab12d6c1..66ac219bc 100644
--- a/lamindb/__init__.py
+++ b/lamindb/__init__.py
@@ -107,6 +107,14 @@
 
 """
 
+from lamindb_setup.lazy_import import enable_lazy_imports
+from lamindb_setup._check_setup import _check_instance_setup
+
+enable_lazy_imports(
+    "lamindb.models",
+    on_load=lambda mod, attr: _check_instance_setup(from_module="lamindb"),
+)
+
 # ruff: noqa: I001
 # denote a release candidate for 0.1.0 with 0.1rc1, 0.1a1, 0.1b1, etc.
 __version__ = "1.6.2"
@@ -118,55 +126,47 @@
     "ignore", message="The legacy Dask DataFrame implementation is deprecated"
 )
 
-from lamindb_setup._check_setup import InstanceNotSetupError as _InstanceNotSetupError
-from lamindb_setup._check_setup import _check_instance_setup
 from lamindb_setup._connect_instance import connect
 from lamindb_setup.core.upath import UPath
-
 from . import base, errors, setup
 
-
-def __getattr__(name):
-    raise _InstanceNotSetupError()
-
-
-if _check_instance_setup(from_module="lamindb"):
-    del __getattr__  # so that imports work out
-    from . import base
-    from ._tracked import tracked
-    from ._view import view
-    from .core._context import context
-    from .core._settings import settings
-    from .curators._legacy import CatManager as Curator
-    from .models import (
-        Artifact,
-        Collection,
-        Feature,
-        FeatureSet,  # backward compat
-        Person,
-        Project,
-        Reference,
-        Run,
-        Schema,
-        Storage,
-        Transform,
-        ULabel,
-        User,
-        Space,
-        Branch,
-        Record,
-        Sheet,
-    )
-    from .models.save import save
-    from . import core
-    from . import integrations
-    from . import curators
-    from . import examples
-
-    track = context._track
-    finish = context._finish
-    settings.__doc__ = """Global live settings (:class:`~lamindb.core.Settings`)."""
-    context.__doc__ = """Global run context (:class:`~lamindb.core.Context`)."""
-    from django.db.models import Q
-
-    Param = Feature  # backward compat
+# this is just used for auto connect on import now
+_check_instance_setup(from_module="lamindb")
+
+from ._tracked import tracked
+from ._view import view
+from .core._context import context
+from .core._settings import settings
+from .curators._legacy import CatManager as Curator
+from .models import (
+    Artifact,
+    Collection,
+    Feature,
+    FeatureSet,  # backward compat
+    Person,
+    Project,
+    Reference,
+    Run,
+    Schema,
+    Storage,
+    Transform,
+    ULabel,
+    User,
+    Space,
+    Branch,
+    Record,
+    Sheet,
+)
+from .models.save import save
+from . import core
+from . import integrations
+from . import curators
+from . import examples
+
+track = context._track
+finish = context._finish
+settings.__doc__ = """Global live settings (:class:`~lamindb.core.Settings`)."""
+context.__doc__ = """Global run context (:class:`~lamindb.core.Context`)."""
+from django.db.models import Q
+
+Param = Feature  # backward compat
diff --git a/lamindb/curators/_legacy.py b/lamindb/curators/_legacy.py
index 4c65b6f49..b858b3479 100644
--- a/lamindb/curators/_legacy.py
+++ b/lamindb/curators/_legacy.py
@@ -185,12 +185,14 @@ class DataFrameCatManager(CatManager):
     def __init__(
         self,
         df: pd.DataFrame | Artifact,
-        columns_field: FieldAttr = Feature.name,
+        columns_field: FieldAttr = ...,
         columns_names: Iterable[str] | None = None,
         categoricals: dict[str, FieldAttr] | None = None,
         sources: dict[str, SQLRecord] | None = None,
         index: Feature | None = None,
     ) -> None:
+        if columns_names is Ellipsis:
+            columns_names = Feature.name
         self._non_validated = None
         self._index = index
         super().__init__(
@@ -326,9 +328,11 @@ def __init__(
         data: ad.AnnData | Artifact,
         var_index: FieldAttr | None = None,
         categoricals: dict[str, FieldAttr] | None = None,
-        obs_columns: FieldAttr = Feature.name,
+        obs_columns: FieldAttr = ...,
         sources: dict[str, SQLRecord] | None = None,
     ) -> None:
+        if obs_columns is Ellipsis:
+            obs_columns = Feature.name
         if isinstance(var_index, str):
             raise TypeError(
                 "var_index parameter has to be a field, e.g. Gene.ensembl_gene_id"
@@ -932,9 +936,11 @@ def __init__(
         experiment_uri: UPathStr | Artifact,
         var_index: dict[str, tuple[str, FieldAttr]],
         categoricals: dict[str, FieldAttr] | None = None,
-        obs_columns: FieldAttr = Feature.name,
+        obs_columns: FieldAttr = ...,
         sources: dict[str, SQLRecord] | None = None,
     ):
+        if obs_columns is Ellipsis:
+            obs_columns = Feature.name
         self._obs_fields = categoricals or {}
         self._var_fields = var_index
         self._columns_field = obs_columns
@@ -1948,9 +1954,11 @@ def from_df(
     cls,
     df: pd.DataFrame,
     categoricals: dict[str, FieldAttr] | None = None,
-    columns: FieldAttr = Feature.name,
+    columns: FieldAttr = ...,
     organism: str | None = None,
 ) -> DataFrameCatManager:
+    if columns is Ellipsis:
+        columns = Feature.name
     if organism is not None:
         logger.warning("organism is ignored, define it on the dtype level")
     return DataFrameCatManager(
@@ -1966,10 +1974,12 @@ def from_anndata(
     data: ad.AnnData | UPathStr,
     var_index: FieldAttr,
     categoricals: dict[str, FieldAttr] | None = None,
-    obs_columns: FieldAttr = Feature.name,
+    obs_columns: FieldAttr = ...,
     organism: str | None = None,
     sources: dict[str, SQLRecord] | None = None,
 ) -> AnnDataCatManager:
+    if obs_columns is Ellipsis:
+        obs_columns = Feature.name
     if organism is not None:
         logger.warning("organism is ignored, define it on the dtype level")
     return AnnDataCatManager(
@@ -2006,10 +2016,12 @@ def from_tiledbsoma(
     experiment_uri: UPathStr,
     var_index: dict[str, tuple[str, FieldAttr]],
     categoricals: dict[str, FieldAttr] | None = None,
-    obs_columns: FieldAttr = Feature.name,
+    obs_columns: FieldAttr = ...,
     organism: str | None = None,
     sources: dict[str, SQLRecord] | None = None,
 ) -> TiledbsomaCatManager:
+    if obs_columns is Ellipsis:
+        obs_columns = Feature.name
     if organism is not None:
         logger.warning("organism is ignored, define it on the dtype level")
     return TiledbsomaCatManager(
diff --git a/lamindb/curators/core.py b/lamindb/curators/core.py
index e923890cb..23d4ea46d 100644
--- a/lamindb/curators/core.py
+++ b/lamindb/curators/core.py
@@ -1347,7 +1347,7 @@ class DataFrameCatManager:
     def __init__(
         self,
         df: pd.DataFrame | Artifact,
-        columns_field: FieldAttr = Feature.name,
+        columns_field: FieldAttr = ...,
         columns_names: Iterable[str] | None = None,
         categoricals: list[Feature] | None = None,
         sources: dict[str, SQLRecord] | None = None,
@@ -1355,6 +1355,8 @@ def __init__(
         slot: str | None = None,
         maximal_set: bool = False,
     ) -> None:
+        if columns_names is Ellipsis:
+            columns_names = Feature.name
         self._non_validated = None
         self._index = index
         self._artifact: Artifact = None  # pass the dataset as an artifact
diff --git a/sub/bionty b/sub/bionty
index 9063380ae..039720c0f 160000
--- a/sub/bionty
+++ b/sub/bionty
@@ -1 +1 @@
-Subproject commit 9063380ae6a876cb260df0ceeb9f275e94d8222b
+Subproject commit 039720c0f2c780e607b206b0c3874cf10dccd67d
diff --git a/sub/lamindb-setup b/sub/lamindb-setup
index 55fdd458f..a408b8cf4 160000
--- a/sub/lamindb-setup
+++ b/sub/lamindb-setup
@@ -1 +1 @@
-Subproject commit 55fdd458fc85e032034a8ed5804140b75e297408
+Subproject commit a408b8cf4a62cca9634e3a3707f27b711027568b