diff --git a/src/spikeinterface/core/base.py b/src/spikeinterface/core/base.py
index 9dd5d4d5b5..ae563c17ae 100644
--- a/src/spikeinterface/core/base.py
+++ b/src/spikeinterface/core/base.py
@@ -81,6 +81,10 @@ def __init__(self, main_ids: Sequence) -> None:
         # "main_ids" will either be channel_ids or units_ids
         # They are used for properties
         self._main_ids = np.array(main_ids)
+        if self._main_ids.dtype.kind == "T":
+            # numpy's variable-width StringDType, which is what a zarr v3 store hands back for a
+            # string column. Store it as fixed-width unicode like every other source.
+            self._main_ids = np.array(self._main_ids.tolist())
         if len(self._main_ids) > 0:
             assert (
                 self._main_ids.dtype.kind in "uiSU"
diff --git a/src/spikeinterface/core/zarrextractors.py b/src/spikeinterface/core/zarrextractors.py
index 02ae5e9ecd..6cdc1c9fde 100644
--- a/src/spikeinterface/core/zarrextractors.py
+++ b/src/spikeinterface/core/zarrextractors.py
@@ -37,7 +37,7 @@ def super_zarr_open(folder_path: str | Path, mode: str = "r", storage_options: d
 
     Returns
     -------
-    root: zarr.hierarchy.Group
+    root: zarr.Group
         The zarr root group object
 
     Raises
@@ -313,7 +313,7 @@ def __init__(self, spikes_group, segment_slices: np.ndarray):
         self._sample_index = spikes_group["sample_index"]
         self._unit_index = spikes_group["unit_index"]
         self._segment_slices = np.asarray(segment_slices, dtype="int64")
-        self._n = len(self._sample_index)
+        self._n = self._sample_index.shape[0]
         self.dtype = np.dtype(minimum_spike_dtype)
 
     @property
@@ -436,7 +436,7 @@ def __init__(
             spikes = ZarrSpikeVector(spikes_group, segment_slices_list)
         else:
             # Materialize the spike vector in memory and sort it by (segment_index, sample_index, unit_index)
-            spikes = np.zeros(len(spikes_group["sample_index"]), dtype=minimum_spike_dtype)
+            spikes = np.zeros(spikes_group["sample_index"].shape[0], dtype=minimum_spike_dtype)
             spikes["sample_index"] = spikes_group["sample_index"][:]
             spikes["unit_index"] = spikes_group["unit_index"][:]
             for i, (start, end) in enumerate(segment_slices_list):
@@ -610,7 +610,7 @@ def get_default_zarr_compressor(clevel: int = 5):
     return Blosc(cname="zstd", clevel=clevel, shuffle=Blosc.BITSHUFFLE)
 
 
-def add_properties_and_annotations(zarr_group: zarr.hierarchy.Group, recording_or_sorting: BaseRecording | BaseSorting):
+def add_properties_and_annotations(zarr_group: zarr.Group, recording_or_sorting: BaseRecording | BaseSorting):
     # save properties
     prop_group = zarr_group.create_group("properties")
     for key in recording_or_sorting.get_property_keys():
@@ -624,7 +624,7 @@ def add_properties_and_annotations(zarr_group: zarr.hierarchy.Group, recording_o
     zarr_group.attrs["annotations"] = check_json(recording_or_sorting._annotations)
 
 
-def add_sorting_to_zarr_group(sorting: BaseSorting, zarr_group: zarr.hierarchy.Group, **kwargs):
+def add_sorting_to_zarr_group(sorting: BaseSorting, zarr_group: zarr.Group, **kwargs):
     """
     Add a sorting extractor to a zarr group.
 
@@ -632,7 +632,7 @@ def add_sorting_to_zarr_group(sorting: BaseSorting, zarr_group: zarr.hierarchy.G
     ----------
     sorting : BaseSorting
         The sorting extractor object to be added to the zarr group
-    zarr_group : zarr.hierarchy.Group
+    zarr_group : zarr.Group
         The zarr group
     kwargs : dict
         Other arguments passed to the zarr compressor
@@ -668,9 +668,7 @@ def add_sorting_to_zarr_group(sorting: BaseSorting, zarr_group: zarr.hierarchy.G
 
 
 # Recording
-def add_recording_to_zarr_group(
-    recording: BaseRecording, zarr_group: zarr.hierarchy.Group, verbose=False, dtype=None, **kwargs
-):
+def add_recording_to_zarr_group(recording: BaseRecording, zarr_group: zarr.Group, verbose=False, dtype=None, **kwargs):
     zarr_kwargs, job_kwargs = split_job_kwargs(kwargs)
 
     if recording.check_serializability("json"):
diff --git a/src/spikeinterface/extractors/nwbextractors.py b/src/spikeinterface/extractors/nwbextractors.py
index fd45aa07c2..e6a30e9fb0 100644
--- a/src/spikeinterface/extractors/nwbextractors.py
+++ b/src/spikeinterface/extractors/nwbextractors.py
@@ -305,8 +305,9 @@ def _get_backend_from_local_file(file_path: str | Path) -> str:
         try:
             import zarr
 
-            with zarr.open(file_path, "r") as f:
-                backend = "zarr"
+            # `mode` is keyword-only in zarr v3, and its groups are not context managers
+            zarr.open(file_path, mode="r")
+            backend = "zarr"
         except:
             raise RuntimeError(f"{file_path} is not a valid Zarr folder!")
     else: