opik 1.9.5__py3-none-any.whl → 1.9.39__py3-none-any.whl

opik/rest_api/datasets/client.py

@@ -5,14 +5,22 @@ import typing
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from ..types.dataset_expansion_response import DatasetExpansionResponse
+from ..types.dataset_item_filter import DatasetItemFilter
 from ..types.dataset_item_page_compare import DatasetItemPageCompare
 from ..types.dataset_item_page_public import DatasetItemPagePublic
 from ..types.dataset_item_public import DatasetItemPublic
+from ..types.dataset_item_update import DatasetItemUpdate
 from ..types.dataset_item_write import DatasetItemWrite
+from ..types.dataset_item_write_source import DatasetItemWriteSource
 from ..types.dataset_page_public import DatasetPagePublic
 from ..types.dataset_public import DatasetPublic
+from ..types.dataset_version_diff import DatasetVersionDiff
+from ..types.dataset_version_page_public import DatasetVersionPagePublic
+from ..types.dataset_version_public import DatasetVersionPublic
+from ..types.json_node import JsonNode
 from ..types.page_columns import PageColumns
 from ..types.project_stats_public import ProjectStatsPublic
+from ..types.span_enrichment_options import SpanEnrichmentOptions
 from ..types.trace_enrichment_options import TraceEnrichmentOptions
 from .raw_client import AsyncRawDatasetsClient, RawDatasetsClient
 from .types.dataset_update_visibility import DatasetUpdateVisibility
@@ -37,6 +45,49 @@ class DatasetsClient:
         """
         return self._raw_client
 
+    def batch_update_dataset_items(
+        self,
+        *,
+        update: DatasetItemUpdate,
+        ids: typing.Optional[typing.Sequence[str]] = OMIT,
+        filters: typing.Optional[typing.Sequence[DatasetItemFilter]] = OMIT,
+        merge_tags: typing.Optional[bool] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Update multiple dataset items
+
+        Parameters
+        ----------
+        update : DatasetItemUpdate
+
+        ids : typing.Optional[typing.Sequence[str]]
+            List of dataset item IDs to update (max 1000). Mutually exclusive with 'filters'.
+
+        filters : typing.Optional[typing.Sequence[DatasetItemFilter]]
+
+        merge_tags : typing.Optional[bool]
+            If true, merge tags with existing tags instead of replacing them. Default: false. When using 'filters', this is automatically set to true.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import OpikApi
+        from Opik import DatasetItemUpdate
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.batch_update_dataset_items(update=DatasetItemUpdate(), )
+        """
+        _response = self._raw_client.batch_update_dataset_items(
+            update=update, ids=ids, filters=filters, merge_tags=merge_tags, request_options=request_options
+        )
+        return _response.data
+
     def find_datasets(
         self,
         *,
@@ -182,6 +233,80 @@ class DatasetsClient:
         )
         return _response.data
 
+    def create_dataset_items_from_csv(
+        self,
+        *,
+        file: typing.Dict[str, typing.Optional[typing.Any]],
+        dataset_id: str,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create dataset items from uploaded CSV file. CSV should have headers in the first row. Processing happens asynchronously in batches.
+
+        Parameters
+        ----------
+        file : typing.Dict[str, typing.Optional[typing.Any]]
+
+        dataset_id : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.create_dataset_items_from_csv(file={'key': 'value'
+        }, dataset_id='dataset_id', )
+        """
+        _response = self._raw_client.create_dataset_items_from_csv(
+            file=file, dataset_id=dataset_id, request_options=request_options
+        )
+        return _response.data
+
+    def create_dataset_items_from_spans(
+        self,
+        dataset_id: str,
+        *,
+        span_ids: typing.Sequence[str],
+        enrichment_options: SpanEnrichmentOptions,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create dataset items from spans with enriched metadata
+
+        Parameters
+        ----------
+        dataset_id : str
+
+        span_ids : typing.Sequence[str]
+            Set of span IDs to add to the dataset
+
+        enrichment_options : SpanEnrichmentOptions
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import OpikApi
+        from Opik import SpanEnrichmentOptions
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.create_dataset_items_from_spans(dataset_id='dataset_id', span_ids=['span_ids'], enrichment_options=SpanEnrichmentOptions(), )
+        """
+        _response = self._raw_client.create_dataset_items_from_spans(
+            dataset_id, span_ids=span_ids, enrichment_options=enrichment_options, request_options=request_options
+        )
+        return _response.data
+
     def create_dataset_items_from_traces(
         self,
         dataset_id: str,
@@ -604,12 +729,70 @@ class DatasetsClient:
         _response = self._raw_client.get_dataset_item_by_id(item_id, request_options=request_options)
         return _response.data
 
+    def patch_dataset_item(
+        self,
+        item_id: str,
+        *,
+        source: DatasetItemWriteSource,
+        data: JsonNode,
+        id: typing.Optional[str] = OMIT,
+        trace_id: typing.Optional[str] = OMIT,
+        span_id: typing.Optional[str] = OMIT,
+        tags: typing.Optional[typing.Sequence[str]] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Partially update dataset item by id. Only provided fields will be updated.
+
+        Parameters
+        ----------
+        item_id : str
+
+        source : DatasetItemWriteSource
+
+        data : JsonNode
+
+        id : typing.Optional[str]
+
+        trace_id : typing.Optional[str]
+
+        span_id : typing.Optional[str]
+
+        tags : typing.Optional[typing.Sequence[str]]
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.patch_dataset_item(item_id='itemId', source="manual", data={'key': 'value'
+        }, )
+        """
+        _response = self._raw_client.patch_dataset_item(
+            item_id,
+            source=source,
+            data=data,
+            id=id,
+            trace_id=trace_id,
+            span_id=span_id,
+            tags=tags,
+            request_options=request_options,
+        )
+        return _response.data
+
     def get_dataset_items(
         self,
         id: str,
         *,
         page: typing.Optional[int] = None,
         size: typing.Optional[int] = None,
+        version: typing.Optional[str] = None,
         filters: typing.Optional[str] = None,
         truncate: typing.Optional[bool] = None,
         request_options: typing.Optional[RequestOptions] = None,
@@ -625,6 +808,8 @@ class DatasetsClient:
 
         size : typing.Optional[int]
 
+        version : typing.Optional[str]
+
         filters : typing.Optional[str]
 
         truncate : typing.Optional[bool]
@@ -644,7 +829,13 @@ class DatasetsClient:
         client.datasets.get_dataset_items(id='id', )
         """
         _response = self._raw_client.get_dataset_items(
-            id, page=page, size=size, filters=filters, truncate=truncate, request_options=request_options
+            id,
+            page=page,
+            size=size,
+            version=version,
+            filters=filters,
+            truncate=truncate,
+            request_options=request_options,
         )
         return _response.data
 
@@ -718,110 +909,123 @@ class DatasetsClient:
         ) as r:
             yield from r.data
 
+    def compare_dataset_versions(
+        self, id: str, *, request_options: typing.Optional[RequestOptions] = None
+    ) -> DatasetVersionDiff:
+        """
+        Compare the latest committed dataset version with the current draft state. This endpoint provides insights into changes made since the last version was committed. The comparison calculates additions, modifications, deletions, and unchanged items between the latest version snapshot and current draft.
 
-class AsyncDatasetsClient:
-    def __init__(self, *, client_wrapper: AsyncClientWrapper):
-        self._raw_client = AsyncRawDatasetsClient(client_wrapper=client_wrapper)
+        Parameters
+        ----------
+        id : str
 
-    @property
-    def with_raw_response(self) -> AsyncRawDatasetsClient:
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetVersionDiff
+            Diff computed successfully
+
+        Examples
+        --------
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.compare_dataset_versions(id='id', )
         """
-        Retrieves a raw implementation of this client that returns raw responses.
+        _response = self._raw_client.compare_dataset_versions(id, request_options=request_options)
+        return _response.data
+
+    def create_version_tag(
+        self, version_hash: str, id: str, *, tag: str, request_options: typing.Optional[RequestOptions] = None
+    ) -> None:
+        """
+        Add a tag to a specific dataset version for easy reference (e.g., 'baseline', 'v1.0', 'production')
+
+        Parameters
+        ----------
+        version_hash : str
+
+        id : str
+
+        tag : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
 
         Returns
         -------
-        AsyncRawDatasetsClient
+        None
+
+        Examples
+        --------
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.create_version_tag(version_hash='versionHash', id='id', tag='tag', )
         """
-        return self._raw_client
+        _response = self._raw_client.create_version_tag(version_hash, id, tag=tag, request_options=request_options)
+        return _response.data
 
-    async def find_datasets(
+    def list_dataset_versions(
         self,
+        id: str,
         *,
         page: typing.Optional[int] = None,
         size: typing.Optional[int] = None,
-        with_experiments_only: typing.Optional[bool] = None,
-        with_optimizations_only: typing.Optional[bool] = None,
-        prompt_id: typing.Optional[str] = None,
-        name: typing.Optional[str] = None,
-        sorting: typing.Optional[str] = None,
-        filters: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
-    ) -> DatasetPagePublic:
+    ) -> DatasetVersionPagePublic:
         """
-        Find datasets
+        Get paginated list of versions for a dataset, ordered by creation time (newest first)
 
         Parameters
         ----------
+        id : str
+
         page : typing.Optional[int]
 
         size : typing.Optional[int]
 
-        with_experiments_only : typing.Optional[bool]
-
-        with_optimizations_only : typing.Optional[bool]
-
-        prompt_id : typing.Optional[str]
-
-        name : typing.Optional[str]
-
-        sorting : typing.Optional[str]
-
-        filters : typing.Optional[str]
-
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
 
         Returns
         -------
-        DatasetPagePublic
-            Dataset resource
+        DatasetVersionPagePublic
+            Dataset versions
 
         Examples
         --------
-        from Opik import AsyncOpikApi
-        import asyncio
-        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
-        async def main() -> None:
-            await client.datasets.find_datasets()
-        asyncio.run(main())
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.list_dataset_versions(id='id', )
         """
-        _response = await self._raw_client.find_datasets(
-            page=page,
-            size=size,
-            with_experiments_only=with_experiments_only,
-            with_optimizations_only=with_optimizations_only,
-            prompt_id=prompt_id,
-            name=name,
-            sorting=sorting,
-            filters=filters,
-            request_options=request_options,
-        )
+        _response = self._raw_client.list_dataset_versions(id, page=page, size=size, request_options=request_options)
         return _response.data
 
-    async def create_dataset(
+    def create_dataset_version(
         self,
+        id: str,
         *,
-        name: str,
-        id: typing.Optional[str] = OMIT,
-        visibility: typing.Optional[DatasetWriteVisibility] = OMIT,
-        tags: typing.Optional[typing.Sequence[str]] = OMIT,
-        description: typing.Optional[str] = OMIT,
+        tag: typing.Optional[str] = OMIT,
+        change_description: typing.Optional[str] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, str]] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> None:
         """
-        Create dataset
+        Create a new immutable version of the dataset by snapshotting the current state
 
         Parameters
         ----------
-        name : str
-
-        id : typing.Optional[str]
+        id : str
 
-        visibility : typing.Optional[DatasetWriteVisibility]
+        tag : typing.Optional[str]
+            Optional tag for this version
 
-        tags : typing.Optional[typing.Sequence[str]]
+        change_description : typing.Optional[str]
+            Optional description of changes in this version
 
-        description : typing.Optional[str]
+        metadata : typing.Optional[typing.Dict[str, str]]
+            Optional user-defined metadata
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -832,38 +1036,28 @@ class AsyncDatasetsClient:
 
         Examples
         --------
-        from Opik import AsyncOpikApi
-        import asyncio
-        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
-        async def main() -> None:
-            await client.datasets.create_dataset(name='name', )
-        asyncio.run(main())
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.create_dataset_version(id='id', )
         """
-        _response = await self._raw_client.create_dataset(
-            name=name, id=id, visibility=visibility, tags=tags, description=description, request_options=request_options
+        _response = self._raw_client.create_dataset_version(
+            id, tag=tag, change_description=change_description, metadata=metadata, request_options=request_options
         )
         return _response.data
 
-    async def create_or_update_dataset_items(
-        self,
-        *,
-        items: typing.Sequence[DatasetItemWrite],
-        dataset_name: typing.Optional[str] = OMIT,
-        dataset_id: typing.Optional[str] = OMIT,
-        request_options: typing.Optional[RequestOptions] = None,
+    def delete_version_tag(
+        self, version_hash: str, tag: str, id: str, *, request_options: typing.Optional[RequestOptions] = None
     ) -> None:
         """
-        Create/update dataset items based on dataset item id
+        Remove a tag from a dataset version. The version itself is not deleted, only the tag reference.
 
         Parameters
         ----------
-        items : typing.Sequence[DatasetItemWrite]
+        version_hash : str
 
-        dataset_name : typing.Optional[str]
-            If null, dataset_id must be provided
+        tag : str
 
-        dataset_id : typing.Optional[str]
-            If null, dataset_name must be provided
+        id : str
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -874,11 +1068,252 @@ class AsyncDatasetsClient:
 
         Examples
         --------
-        from Opik import AsyncOpikApi
-        from Opik import DatasetItemWrite
-        import asyncio
-        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
-        async def main() -> None:
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.delete_version_tag(version_hash='versionHash', tag='tag', id='id', )
+        """
+        _response = self._raw_client.delete_version_tag(version_hash, tag, id, request_options=request_options)
+        return _response.data
+
+    def restore_dataset_version(
+        self, id: str, *, version_ref: str, request_options: typing.Optional[RequestOptions] = None
+    ) -> DatasetVersionPublic:
+        """
+        Restores the dataset to a previous version state. All draft items are replaced with items from the specified version. If the version is not the latest, a new version snapshot is created. If the version is the latest, only draft items are replaced (revert functionality).
+
+        Parameters
+        ----------
+        id : str
+
+        version_ref : str
+            Version hash or tag to restore from
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetVersionPublic
+            Version restored successfully
+
+        Examples
+        --------
+        from Opik import OpikApi
+        client = OpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        client.datasets.restore_dataset_version(id='id', version_ref='version_ref', )
+        """
+        _response = self._raw_client.restore_dataset_version(
+            id, version_ref=version_ref, request_options=request_options
+        )
+        return _response.data
+
+
+class AsyncDatasetsClient:
+    def __init__(self, *, client_wrapper: AsyncClientWrapper):
+        self._raw_client = AsyncRawDatasetsClient(client_wrapper=client_wrapper)
+
+    @property
+    def with_raw_response(self) -> AsyncRawDatasetsClient:
+        """
+        Retrieves a raw implementation of this client that returns raw responses.
+
+        Returns
+        -------
+        AsyncRawDatasetsClient
+        """
+        return self._raw_client
+
+    async def batch_update_dataset_items(
+        self,
+        *,
+        update: DatasetItemUpdate,
+        ids: typing.Optional[typing.Sequence[str]] = OMIT,
+        filters: typing.Optional[typing.Sequence[DatasetItemFilter]] = OMIT,
+        merge_tags: typing.Optional[bool] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Update multiple dataset items
+
+        Parameters
+        ----------
+        update : DatasetItemUpdate
+
+        ids : typing.Optional[typing.Sequence[str]]
+            List of dataset item IDs to update (max 1000). Mutually exclusive with 'filters'.
+
+        filters : typing.Optional[typing.Sequence[DatasetItemFilter]]
+
+        merge_tags : typing.Optional[bool]
+            If true, merge tags with existing tags instead of replacing them. Default: false. When using 'filters', this is automatically set to true.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        from Opik import DatasetItemUpdate
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.batch_update_dataset_items(update=DatasetItemUpdate(), )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.batch_update_dataset_items(
+            update=update, ids=ids, filters=filters, merge_tags=merge_tags, request_options=request_options
+        )
+        return _response.data
+
+    async def find_datasets(
+        self,
+        *,
+        page: typing.Optional[int] = None,
+        size: typing.Optional[int] = None,
+        with_experiments_only: typing.Optional[bool] = None,
+        with_optimizations_only: typing.Optional[bool] = None,
+        prompt_id: typing.Optional[str] = None,
+        name: typing.Optional[str] = None,
+        sorting: typing.Optional[str] = None,
+        filters: typing.Optional[str] = None,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> DatasetPagePublic:
+        """
+        Find datasets
+
+        Parameters
+        ----------
+        page : typing.Optional[int]
+
+        size : typing.Optional[int]
+
+        with_experiments_only : typing.Optional[bool]
+
+        with_optimizations_only : typing.Optional[bool]
+
+        prompt_id : typing.Optional[str]
+
+        name : typing.Optional[str]
+
+        sorting : typing.Optional[str]
+
+        filters : typing.Optional[str]
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetPagePublic
+            Dataset resource
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.find_datasets()
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.find_datasets(
+            page=page,
+            size=size,
+            with_experiments_only=with_experiments_only,
+            with_optimizations_only=with_optimizations_only,
+            prompt_id=prompt_id,
+            name=name,
+            sorting=sorting,
+            filters=filters,
+            request_options=request_options,
+        )
+        return _response.data
+
+    async def create_dataset(
+        self,
+        *,
+        name: str,
+        id: typing.Optional[str] = OMIT,
+        visibility: typing.Optional[DatasetWriteVisibility] = OMIT,
+        tags: typing.Optional[typing.Sequence[str]] = OMIT,
+        description: typing.Optional[str] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create dataset
+
+        Parameters
+        ----------
+        name : str
+
+        id : typing.Optional[str]
+
+        visibility : typing.Optional[DatasetWriteVisibility]
+
+        tags : typing.Optional[typing.Sequence[str]]
+
+        description : typing.Optional[str]
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.create_dataset(name='name', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.create_dataset(
+            name=name, id=id, visibility=visibility, tags=tags, description=description, request_options=request_options
+        )
+        return _response.data
+
+    async def create_or_update_dataset_items(
+        self,
+        *,
+        items: typing.Sequence[DatasetItemWrite],
+        dataset_name: typing.Optional[str] = OMIT,
+        dataset_id: typing.Optional[str] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create/update dataset items based on dataset item id
+
+        Parameters
+        ----------
+        items : typing.Sequence[DatasetItemWrite]
+
+        dataset_name : typing.Optional[str]
+            If null, dataset_id must be provided
+
+        dataset_id : typing.Optional[str]
+            If null, dataset_name must be provided
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        from Opik import DatasetItemWrite
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
             await client.datasets.create_or_update_dataset_items(items=[DatasetItemWrite(source="manual", data={'key': 'value'
             }, )], )
         asyncio.run(main())
@@ -888,6 +1323,86 @@ class AsyncDatasetsClient:
         )
         return _response.data
 
+    async def create_dataset_items_from_csv(
+        self,
+        *,
+        file: typing.Dict[str, typing.Optional[typing.Any]],
+        dataset_id: str,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create dataset items from uploaded CSV file. CSV should have headers in the first row. Processing happens asynchronously in batches.
+
+        Parameters
+        ----------
+        file : typing.Dict[str, typing.Optional[typing.Any]]
+
+        dataset_id : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.create_dataset_items_from_csv(file={'key': 'value'
+            }, dataset_id='dataset_id', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.create_dataset_items_from_csv(
+            file=file, dataset_id=dataset_id, request_options=request_options
+        )
+        return _response.data
+
+    async def create_dataset_items_from_spans(
+        self,
+        dataset_id: str,
+        *,
+        span_ids: typing.Sequence[str],
+        enrichment_options: SpanEnrichmentOptions,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create dataset items from spans with enriched metadata
+
+        Parameters
+        ----------
+        dataset_id : str
+
+        span_ids : typing.Sequence[str]
+            Set of span IDs to add to the dataset
+
+        enrichment_options : SpanEnrichmentOptions
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        from Opik import SpanEnrichmentOptions
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.create_dataset_items_from_spans(dataset_id='dataset_id', span_ids=['span_ids'], enrichment_options=SpanEnrichmentOptions(), )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.create_dataset_items_from_spans(
+            dataset_id, span_ids=span_ids, enrichment_options=enrichment_options, request_options=request_options
+        )
+        return _response.data
+
     async def create_dataset_items_from_traces(
         self,
         dataset_id: str,
@@ -1350,12 +1865,73 @@ class AsyncDatasetsClient:
         _response = await self._raw_client.get_dataset_item_by_id(item_id, request_options=request_options)
         return _response.data
 
+    async def patch_dataset_item(
+        self,
+        item_id: str,
+        *,
+        source: DatasetItemWriteSource,
+        data: JsonNode,
+        id: typing.Optional[str] = OMIT,
+        trace_id: typing.Optional[str] = OMIT,
+        span_id: typing.Optional[str] = OMIT,
+        tags: typing.Optional[typing.Sequence[str]] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Partially update dataset item by id. Only provided fields will be updated.
+
+        Parameters
+        ----------
+        item_id : str
+
+        source : DatasetItemWriteSource
+
+        data : JsonNode
+
+        id : typing.Optional[str]
+
+        trace_id : typing.Optional[str]
+
+        span_id : typing.Optional[str]
+
+        tags : typing.Optional[typing.Sequence[str]]
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.patch_dataset_item(item_id='itemId', source="manual", data={'key': 'value'
+            }, )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.patch_dataset_item(
+            item_id,
+            source=source,
+            data=data,
+            id=id,
+            trace_id=trace_id,
+            span_id=span_id,
+            tags=tags,
+            request_options=request_options,
+        )
+        return _response.data
+
     async def get_dataset_items(
         self,
         id: str,
         *,
         page: typing.Optional[int] = None,
         size: typing.Optional[int] = None,
+        version: typing.Optional[str] = None,
         filters: typing.Optional[str] = None,
         truncate: typing.Optional[bool] = None,
         request_options: typing.Optional[RequestOptions] = None,
@@ -1371,6 +1947,8 @@ class AsyncDatasetsClient:
 
         size : typing.Optional[int]
 
+        version : typing.Optional[str]
+
         filters : typing.Optional[str]
 
         truncate : typing.Optional[bool]
@@ -1393,7 +1971,13 @@ class AsyncDatasetsClient:
         asyncio.run(main())
         """
         _response = await self._raw_client.get_dataset_items(
-            id, page=page, size=size, filters=filters, truncate=truncate, request_options=request_options
+            id,
+            page=page,
+            size=size,
+            version=version,
+            filters=filters,
+            truncate=truncate,
+            request_options=request_options,
         )
         return _response.data
 
@@ -1470,3 +2054,223 @@ class AsyncDatasetsClient:
         ) as r:
             async for data in r.data:
                 yield data
+
+    async def compare_dataset_versions(
+        self, id: str, *, request_options: typing.Optional[RequestOptions] = None
+    ) -> DatasetVersionDiff:
+        """
+        Compare the latest committed dataset version with the current draft state. This endpoint provides insights into changes made since the last version was committed. The comparison calculates additions, modifications, deletions, and unchanged items between the latest version snapshot and current draft.
+
+        Parameters
+        ----------
+        id : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetVersionDiff
+            Diff computed successfully
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.compare_dataset_versions(id='id', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.compare_dataset_versions(id, request_options=request_options)
+        return _response.data
+
+    async def create_version_tag(
+        self, version_hash: str, id: str, *, tag: str, request_options: typing.Optional[RequestOptions] = None
+    ) -> None:
+        """
+        Add a tag to a specific dataset version for easy reference (e.g., 'baseline', 'v1.0', 'production')
+
+        Parameters
+        ----------
+        version_hash : str
+
+        id : str
+
+        tag : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.create_version_tag(version_hash='versionHash', id='id', tag='tag', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.create_version_tag(
+            version_hash, id, tag=tag, request_options=request_options
+        )
+        return _response.data
+
+    async def list_dataset_versions(
+        self,
+        id: str,
+        *,
+        page: typing.Optional[int] = None,
+        size: typing.Optional[int] = None,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> DatasetVersionPagePublic:
+        """
+        Get paginated list of versions for a dataset, ordered by creation time (newest first)
+
+        Parameters
+        ----------
+        id : str
+
+        page : typing.Optional[int]
+
+        size : typing.Optional[int]
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetVersionPagePublic
+            Dataset versions
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.list_dataset_versions(id='id', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.list_dataset_versions(
+            id, page=page, size=size, request_options=request_options
+        )
+        return _response.data
+
+    async def create_dataset_version(
+        self,
+        id: str,
+        *,
+        tag: typing.Optional[str] = OMIT,
+        change_description: typing.Optional[str] = OMIT,
+        metadata: typing.Optional[typing.Dict[str, str]] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> None:
+        """
+        Create a new immutable version of the dataset by snapshotting the current state
+
+        Parameters
+        ----------
+        id : str
+
+        tag : typing.Optional[str]
+            Optional tag for this version
+
+        change_description : typing.Optional[str]
+            Optional description of changes in this version
+
+        metadata : typing.Optional[typing.Dict[str, str]]
+            Optional user-defined metadata
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.create_dataset_version(id='id', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.create_dataset_version(
+            id, tag=tag, change_description=change_description, metadata=metadata, request_options=request_options
+        )
+        return _response.data
+
+    async def delete_version_tag(
+        self, version_hash: str, tag: str, id: str, *, request_options: typing.Optional[RequestOptions] = None
+    ) -> None:
+        """
+        Remove a tag from a dataset version. The version itself is not deleted, only the tag reference.
+
+        Parameters
+        ----------
+        version_hash : str
+
+        tag : str
+
+        id : str
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        None
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.delete_version_tag(version_hash='versionHash', tag='tag', id='id', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.delete_version_tag(version_hash, tag, id, request_options=request_options)
+        return _response.data
+
+    async def restore_dataset_version(
+        self, id: str, *, version_ref: str, request_options: typing.Optional[RequestOptions] = None
+    ) -> DatasetVersionPublic:
+        """
+        Restores the dataset to a previous version state. All draft items are replaced with items from the specified version. If the version is not the latest, a new version snapshot is created. If the version is the latest, only draft items are replaced (revert functionality).
+
+        Parameters
+        ----------
+        id : str
+
+        version_ref : str
+            Version hash or tag to restore from
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        DatasetVersionPublic
+            Version restored successfully
+
+        Examples
+        --------
+        from Opik import AsyncOpikApi
+        import asyncio
+        client = AsyncOpikApi(api_key="YOUR_API_KEY", workspace_name="YOUR_WORKSPACE_NAME", )
+        async def main() -> None:
+            await client.datasets.restore_dataset_version(id='id', version_ref='version_ref', )
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.restore_dataset_version(
+            id, version_ref=version_ref, request_options=request_options
+        )
+        return _response.data