snowflake-ml-python 1.49.0__tar.gz → 1.50.0__tar.gz

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/CHANGELOG.md

@@ -1,18 +1,28 @@
 # Release History
 
-## 1.49.0
+## 1.50.0
 
 ### New Features
 
 ### Bug Fixes
 
-* Registry: Fixed a bug where logging a model with explainability enabled could fail with a data validation error
-  ("There is no non-null data in column ...") when a numeric input column was entirely null within the background
-  sample. The `explain` method now reuses the model's existing input signature instead of re-inferring it from the
-  sample.
+### Behavior Changes
 
-* Feature Store: Fixed a bug where `get_feature_view()` could return stale metadata after re-registering a feature
-  view with `overwrite=True`. Metadata reads now select the most recently written row.
+### Deprecations
+
+## 1.49.0 (2026-07-29)
+
+### New Features
+
+* Experiment Tracking: `ExperimentTracking` now captures best-effort source provenance (entry-point filename and any
+  surrounding git commit, branch, and remote URL) for each new run by default. Pass `capture_source_info=False` to
+  disable it. Collection is always non-fatal and never blocks run creation.
+
+### Bug Fixes
+
+* Registry: Fixed `ModelVersion.load()` failing with `Result for query <id> has expired` inside Snowflake Container
+  Runtime notebooks. Model file downloads now use the `session.file.get` FileOperation API (the same mechanism used
+  when logging a model) instead of a raw `GET` query, avoiding a dependency on `collect()` result sets.
 
 ### Behavior Changes
 
@@ -28,6 +38,11 @@
 
 ### Bug Fixes
 
+* Registry: Fixed a bug where logging a model with explainability enabled could fail with a data validation error
+  ("There is no non-null data in column ...") when a numeric input column was entirely null within the background
+  sample. The `explain` method now reuses the model's existing input signature instead of re-inferring it from the
+  sample.
+
 ### Behavior Changes
 
 ### Deprecations

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: snowflake-ml-python
-Version: 1.49.0
+Version: 1.50.0
 Summary: The machine learning client library that is used for interacting with Snowflake to build machine learning solutions.
 Author-email: "Snowflake, Inc" <support@snowflake.com>
 License: 
@@ -236,7 +236,6 @@ License-File: LICENSE.txt
 Requires-Dist: anyio<5,>=3.5.0
 Requires-Dist: cachetools>=3.1.1
 Requires-Dist: cloudpickle>=2.0.0
-Requires-Dist: cryptography
 Requires-Dist: fsspec[http]<2026,>=2024.6.1
 Requires-Dist: h2<5,>=4.3.0
 Requires-Dist: importlib_resources<7,>=6.1.1
@@ -417,19 +416,29 @@ NOTE: Version 1.7.0 is used as example here. Please choose the the latest versio
 
 # Release History
 
-## 1.49.0
+## 1.50.0
 
 ### New Features
 
 ### Bug Fixes
 
-* Registry: Fixed a bug where logging a model with explainability enabled could fail with a data validation error
-  ("There is no non-null data in column ...") when a numeric input column was entirely null within the background
-  sample. The `explain` method now reuses the model's existing input signature instead of re-inferring it from the
-  sample.
+### Behavior Changes
 
-* Feature Store: Fixed a bug where `get_feature_view()` could return stale metadata after re-registering a feature
-  view with `overwrite=True`. Metadata reads now select the most recently written row.
+### Deprecations
+
+## 1.49.0 (2026-07-29)
+
+### New Features
+
+* Experiment Tracking: `ExperimentTracking` now captures best-effort source provenance (entry-point filename and any
+  surrounding git commit, branch, and remote URL) for each new run by default. Pass `capture_source_info=False` to
+  disable it. Collection is always non-fatal and never blocks run creation.
+
+### Bug Fixes
+
+* Registry: Fixed `ModelVersion.load()` failing with `Result for query <id> has expired` inside Snowflake Container
+  Runtime notebooks. Model file downloads now use the `session.file.get` FileOperation API (the same mechanism used
+  when logging a model) instead of a raw `GET` query, avoiding a dependency on `collect()` result sets.
 
 ### Behavior Changes
 
@@ -445,6 +454,11 @@ NOTE: Version 1.7.0 is used as example here. Please choose the the latest versio
 
 ### Bug Fixes
 
+* Registry: Fixed a bug where logging a model with explainability enabled could fail with a data validation error
+  ("There is no non-null data in column ...") when a numeric input column was entirely null within the background
+  sample. The `explain` method now reuses the model's existing input signature instead of re-inferring it from the
+  sample.
+
 ### Behavior Changes
 
 ### Deprecations

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/pyproject.toml

@@ -8,7 +8,7 @@ description = "The machine learning client library that is used for interacting
 classifiers = [ "Development Status :: 5 - Production/Stable", "Environment :: Console", "Environment :: Other Environment", "Intended Audience :: Developers", "Intended Audience :: Education", "Intended Audience :: Information Technology", "Intended Audience :: System Administrators", "License :: OSI Approved :: Apache Software License", "Operating System :: OS Independent", "Programming Language :: Python :: 3.9", "Programming Language :: Python :: 3.10", "Programming Language :: Python :: 3.11", "Programming Language :: Python :: 3.12", "Topic :: Database", "Topic :: Software Development", "Topic :: Software Development :: Libraries", "Topic :: Software Development :: Libraries :: Application Frameworks", "Topic :: Software Development :: Libraries :: Python Modules", "Topic :: Scientific/Engineering :: Information Analysis",]
 requires-python = ">=3.9, <3.14"
 dynamic = [ "version", "readme",]
-dependencies = [ "anyio>=3.5.0,<5", "cachetools>=3.1.1", "cloudpickle>=2.0.0", "cryptography", "fsspec[http]>=2024.6.1,<2026", "h2>=4.3.0,<5", "importlib_resources>=6.1.1, <7", "jinja2>=3.1.0,<4", "numpy>=1.23,<3", "packaging>=20.9,<26", "pandas>=2.1.4,<3", "platformdirs<5", "pyarrow", "pydantic>=2.8.2, <3", "pyjwt>=2.0.0, <3", "pytimeparse>=1.1.8,<2", "pyyaml>=6.0,<7", "retrying>=1.3.3,<2", "scikit-learn<1.9", "scipy>=1.9,<2", "shap>=0.46.0,<1", "snowflake-connector-python[pandas]>=3.17.3,<5", "snowflake-snowpark-python>=1.17.0,<2,!=1.26.0", "sqlparse>=0.4,<1", "tqdm<5", "typing-extensions>=4.1.0,<5", "xgboost<4",]
+dependencies = [ "anyio>=3.5.0,<5", "cachetools>=3.1.1", "cloudpickle>=2.0.0", "fsspec[http]>=2024.6.1,<2026", "h2>=4.3.0,<5", "importlib_resources>=6.1.1, <7", "jinja2>=3.1.0,<4", "numpy>=1.23,<3", "packaging>=20.9,<26", "pandas>=2.1.4,<3", "platformdirs<5", "pyarrow", "pydantic>=2.8.2, <3", "pyjwt>=2.0.0, <3", "pytimeparse>=1.1.8,<2", "pyyaml>=6.0,<7", "retrying>=1.3.3,<2", "scikit-learn<1.9", "scipy>=1.9,<2", "shap>=0.46.0,<1", "snowflake-connector-python[pandas]>=3.17.3,<5", "snowflake-snowpark-python>=1.17.0,<2,!=1.26.0", "sqlparse>=0.4,<1", "tqdm<5", "typing-extensions>=4.1.0,<5", "xgboost<4",]
 [[project.authors]]
 name = "Snowflake, Inc"
 email = "support@snowflake.com"

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/_internal/utils/query_result_checker.py

@@ -270,4 +270,6 @@ class SqlResultValidator(ResultValidator):
 
     def _get_result(self) -> list[snowpark.Row]:
         """Collect the result of the given SQL query."""
+        if self._params is None:
+            return self._session.sql(self._query).collect(statement_params=self._statement_params)
         return self._session.sql(self._query, params=self._params).collect(statement_params=self._statement_params)

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/feature_store/feature_store.py

@@ -27,7 +27,6 @@ from typing_extensions import Concatenate, ParamSpec
 
 import snowflake.ml.feature_store.feature_view as fv_mod
 import snowflake.ml.version as snowml_version
-from snowflake.ml import dataset
 from snowflake.ml._internal import telemetry
 from snowflake.ml._internal.exceptions import (
     dataset_errors,
@@ -42,7 +41,10 @@ from snowflake.ml._internal.utils.sql_identifier import (
     parse_fully_qualified_name,
     to_sql_identifiers,
 )
-from snowflake.ml.dataset.dataset_metadata import FeatureStoreMetadata
+
+if TYPE_CHECKING:
+    from snowflake.ml import dataset
+
 from snowflake.ml.feature_store import (
     feature_group as fg_mod,
     feature_view_append_only_validation,
@@ -3783,6 +3785,11 @@ class FeatureStore:
             -------------------------------------------------------
 
         """
+        from snowflake.ml import (
+            dataset,  # lazy import — heavy deps (numpy/pandas/pyarrow)
+        )
+        from snowflake.ml.dataset.dataset_metadata import FeatureStoreMetadata
+
         if output_type not in {"table", "dataset"}:
             raise snowml_exceptions.SnowflakeMLException(
                 error_code=error_codes.INVALID_ARGUMENT,
@@ -3904,6 +3911,10 @@ class FeatureStore:
             1.0
 
         """
+        from snowflake.ml.dataset.dataset_metadata import (
+            FeatureStoreMetadata,  # lazy import
+        )
+
         assert ds.selected_version is not None
         source_meta = ds.selected_version._get_metadata()
         if (

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/feature_store/feature_view.py

@@ -66,6 +66,34 @@ _PG_IDENTIFIER_BYTE_LIMIT = 63
 _POSTGRES_ONLINE_MAX_NAME_VERSION_LEN = 45
 _POSTGRES_ONLINE_MAX_COLUMN_LEN = 30
 
+# The online-store schema is derived as "{schema}__{oft_id}", so its budget reserves room for the "__"
+# separator and a 19-digit (64-bit) oft_id, keeping the derived identifier within the limit (63 - 2 - 19).
+_POSTGRES_ONLINE_MAX_SCHEMA_LEN = 42
+
+
+def validate_postgres_online_schema_length(feature_store_schema: str) -> None:
+    """Reject a Feature Store schema too long for the Postgres online-store schema identifier.
+
+    The online store derives its schema as "{feature_store_schema}__{oft_id}" and must fit PostgreSQL's
+    63-byte identifier limit, so the budget reserves room for the "__" separator and the downstream-assigned
+    oft_id. Called from :func:`online_service.create_online_service`, before any feature view or feature
+    group is materialized.
+
+    Args:
+        feature_store_schema: The Feature Store schema name that will prefix the Postgres online-store
+            schema.
+
+    Raises:
+        ValueError: If the schema length exceeds the Postgres online-store budget.
+    """
+    if len(feature_store_schema) > _POSTGRES_ONLINE_MAX_SCHEMA_LEN:
+        raise ValueError(
+            f"feature store schema '{feature_store_schema}' is too long for the Postgres online store: "
+            f"it uses {len(feature_store_schema)} characters, but the limit is "
+            f"{_POSTGRES_ONLINE_MAX_SCHEMA_LEN}. Use a shorter Feature Store schema."
+        )
+
+
 _RESULT_SCAN_QUERY_PATTERN = re.compile(
     r".*FROM\s*TABLE\s*\(\s*RESULT_SCAN\s*\(.*",
     flags=re.DOTALL | re.IGNORECASE | re.X,

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/feature_store/online_service.py

@@ -241,6 +241,16 @@ def create_online_service(
             error_code=error_codes.INVALID_ARGUMENT,
             original_exception=ValueError("producer_role and consumer_role must be non-empty."),
         )
+    from snowflake.ml.feature_store.feature_view import (
+        validate_postgres_online_schema_length,
+    )
+
+    try:
+        validate_postgres_online_schema_length(schema.resolved())
+    except ValueError as e:
+        raise snowml_exceptions.SnowflakeMLException(
+            error_code=error_codes.INVALID_ARGUMENT, original_exception=e
+        ) from e
     payload = json.dumps(
         {
             "roles": {

snowflake_ml_python-1.50.0/snowflake/ml/model/_client/model/batch_inference_job_specs.py

@@ -0,0 +1,166 @@
+from enum import Enum
+from typing import Any, Optional
+
+from pydantic import BaseModel, ConfigDict, field_serializer, field_validator
+from typing_extensions import TypedDict
+
+from snowflake.ml.model import inference_engine as inference_engine_module
+
+
+class SaveMode(str, Enum):
+    """Save mode options for batch inference output.
+
+    Determines the behavior when files already exist in the output stage location.
+
+    OVERWRITE: Remove existing files and write new results.
+
+    ERROR: Raise an error if files already exist in the output stage location.
+    """
+
+    OVERWRITE = "overwrite"
+    ERROR = "error"
+
+
+class InputFormat(str, Enum):
+    """The format of the input column data."""
+
+    FULL_STAGE_PATH = "full_stage_path"
+
+
+class FileEncoding(str, Enum):
+    """The encoding of the file content that will be passed to the custom model."""
+
+    RAW_BYTES = "raw_bytes"
+    BASE64 = "base64"
+    BASE64_DATA_URL = "base64_data_url"
+
+
+class ColumnHandlingOptions(TypedDict):
+    """Options for handling specific columns during run_batch for file I/O."""
+
+    input_format: InputFormat
+    convert_to: FileEncoding
+
+
+class InputSpec(BaseModel):
+    """Input block of the batch inference job specification.
+
+    Attributes:
+        params (Optional[dict[str, Any]]): Optional dictionary of model inference parameters
+            (e.g., temperature, top_k for LLMs). These are passed as keyword arguments to the
+            model's inference method. Defaults to None.
+        column_handling (Optional[dict[str, ColumnHandlingOptions]]): Optional dictionary
+            specifying how to handle specific columns during file I/O. Maps column names to their
+            input format and file encoding configuration.
+        partition_column (Optional[str]): Optional column name to use for partitioning the input
+            data. When set, the batch inference job will partition the data by this column.
+            Defaults to None.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    params: Optional[dict[str, Any]] = None
+    column_handling: Optional[dict[str, ColumnHandlingOptions]] = None
+    partition_column: Optional[str] = None
+
+
+class OutputSpec(BaseModel):
+    """Output block of the batch inference job specification.
+
+    Results are written under ``<stage_location>/<job_name>/``.
+
+    Attributes:
+        stage_location (str): The stage path under which batch inference results will be saved.
+            This should be a full path including the stage with @ prefix. For example,
+            '@My_DB.PUBLIC.MY_STAGE/some/path/'. Only Snowflake internal stages are supported.
+        mode (SaveMode): The save mode that determines behavior when files already exist
+            at the output stage location. Defaults to SaveMode.ERROR.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    stage_location: str
+    mode: SaveMode = SaveMode.ERROR
+
+
+class ResourcesSpec(BaseModel):
+    """Resources block of the batch inference job specification.
+
+    Attributes:
+        cpu_requests (Optional[str]): The cpu limit for CPU based inference. Can be an integer,
+            fractional or string values. If None, we attempt to utilize all the vCPU of the node.
+        memory_requests (Optional[str]): The memory limit for inference. Can be an integer
+            or a fractional value, but requires a unit (GiB, MiB). If None, we attempt to utilize
+            all the memory of the node.
+        gpu_requests (Optional[str]): The gpu limit for GPU based inference. Can be integer or
+            string values. Use CPU if None.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    cpu_requests: Optional[str] = None
+    memory_requests: Optional[str] = None
+    gpu_requests: Optional[str] = None
+
+
+class EngineOptions(BaseModel):
+    """``inference.engine_options`` sub-block of the job specification.
+
+    Attributes:
+        engine (Optional[InferenceEngine]): The inference engine to use.
+        engine_args_override (Optional[list[str]]): Arguments passed through to the inference engine.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    engine: Optional[inference_engine_module.InferenceEngine] = None
+    engine_args_override: Optional[list[str]] = None
+
+    @field_validator("engine", mode="before")
+    @classmethod
+    def _coerce_engine(cls, value: Any) -> Optional[inference_engine_module.InferenceEngine]:
+        # Accept an InferenceEngine member or a case-insensitive value/name string
+        # (e.g. "vllm", "VLLM", "python_generic") rather than only the exact enum value.
+        if value is None:
+            return None
+        return inference_engine_module.InferenceEngine.from_value(value)
+
+    @field_serializer("engine")
+    def _serialize_engine(self, engine: Optional[inference_engine_module.InferenceEngine]) -> Optional[str]:
+        # Emit the enum member name (e.g. "VLLM") rather than its lower-case value,
+        # to match the canonical spelling used in the EXECUTE INFERENCE JOB SERVICE spec.
+        return engine.name if engine is not None else None
+
+
+class InferenceSpec(BaseModel):
+    """Inference block of the batch inference job specification.
+
+    Attributes:
+        num_workers (Optional[int]): The number of workers to run the inference service for handling
+            requests in parallel within an instance of the service. Auto determined if None.
+        max_batch_rows (Optional[int]): Maximum number of rows to process in a single batch.
+            Auto determined if None. Larger values may improve throughput.
+        engine_options (Optional[EngineOptions]): Options for a custom inference engine.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    num_workers: Optional[int] = None
+    max_batch_rows: Optional[int] = None
+    engine_options: Optional[EngineOptions] = None
+
+
+class ImageBuildSpec(BaseModel):
+    """Image-build block of the batch inference job specification.
+
+    Attributes:
+        image_repo (Optional[str]): Container image repository for the inference job.
+            If not specified, uses the default repository.
+        force_rebuild (bool): Whether to force rebuilding the container image even if
+            it already exists. Defaults to False.
+    """
+
+    model_config = ConfigDict(extra="forbid")
+
+    image_repo: Optional[str] = None
+    force_rebuild: bool = False

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/model/_client/model/batch_inference_serialization.py

@@ -5,7 +5,9 @@ from typing import Any, Optional
 
 from pydantic import TypeAdapter
 
-from snowflake.ml.model._client.model.batch_inference_specs import ColumnHandlingOptions
+from snowflake.ml.model._client.model.batch_inference_job_specs import (
+    ColumnHandlingOptions,
+)
 
 _UTF8_ENCODING = "utf-8"
 

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/model/_client/model/batch_inference_specs.py

@@ -1,45 +1,11 @@
-from enum import Enum
 from typing import Any, Optional
 
-from pydantic import BaseModel, field_serializer, model_validator
-from typing_extensions import TypedDict
+from pydantic import BaseModel, model_validator
 
-from snowflake.ml.model import inference_engine as inference_engine_module
-
-
-class SaveMode(str, Enum):
-    """Save mode options for batch inference output.
-
-    Determines the behavior when files already exist in the output location.
-
-    OVERWRITE: Remove existing files and write new results.
-
-    ERROR: Raise an error if files already exist in the output location.
-    """
-
-    OVERWRITE = "overwrite"
-    ERROR = "error"
-
-
-class InputFormat(str, Enum):
-    """The format of the input column data."""
-
-    FULL_STAGE_PATH = "full_stage_path"
-
-
-class FileEncoding(str, Enum):
-    """The encoding of the file content that will be passed to the custom model."""
-
-    RAW_BYTES = "raw_bytes"
-    BASE64 = "base64"
-    BASE64_DATA_URL = "base64_data_url"
-
-
-class ColumnHandlingOptions(TypedDict):
-    """Options for handling specific columns during run_batch for file I/O."""
-
-    input_format: InputFormat
-    convert_to: FileEncoding
+from snowflake.ml.model._client.model.batch_inference_job_specs import (
+    ColumnHandlingOptions,
+    SaveMode,
+)
 
 
 class InputSpec(BaseModel):
@@ -79,7 +45,7 @@ class OutputSpec(BaseModel):
     """Specification for batch inference output.
 
     Defines where the inference results should be written and how to handle
-    existing files at the output location. Exactly one of ``stage_location`` or
+    existing files at the output stage location. Exactly one of ``stage_location`` or
     ``base_stage_location`` must be provided.
 
     Attributes:
@@ -92,7 +58,7 @@ class OutputSpec(BaseModel):
             derives the output path as ``<base_stage_location>/<job_name>/``.
             Mutually exclusive with stage_location.
         mode (SaveMode): The save mode that determines behavior when files already exist
-            at the output location. Defaults to SaveMode.ERROR which raises an error
+            at the output stage location. Defaults to SaveMode.ERROR which raises an error
             if files exist. Can be set to SaveMode.OVERWRITE to replace existing files.
 
     Example:
@@ -185,67 +151,3 @@ class JobSpec(BaseModel):
         if self.job_name is not None and self.job_name_prefix is not None:
             raise ValueError("job_name and job_name_prefix are mutually exclusive. Please specify only one or neither.")
         return self
-
-
-# ----------------------------------------------------------------------------
-# Spec types for EXECUTE INFERENCE JOB SERVICE.
-#
-# These types mirror the YAML body 1:1 (`input` / `output` / `resources`
-# / `inference` / `image_build`).
-# ----------------------------------------------------------------------------
-
-
-class Input(BaseModel):
-    """Input block of the batch inference YAML body."""
-
-    params: Optional[dict[str, Any]] = None
-    column_handling: Optional[dict[str, ColumnHandlingOptions]] = None
-    partition_column: Optional[str] = None
-
-
-class Output(BaseModel):
-    """Output block of the batch inference YAML body.
-
-    ``stage_location`` is required and is treated as a base: results are written under a
-    per-job subdirectory, ``<stage_location>/<job_name>/``.
-    """
-
-    stage_location: str
-    mode: SaveMode = SaveMode.ERROR
-
-
-class Resources(BaseModel):
-    """Resources block of the batch inference YAML body."""
-
-    cpu_requests: Optional[str] = None
-    memory_requests: Optional[str] = None
-    gpu_requests: Optional[str] = None
-
-
-class EngineOptions(BaseModel):
-    """``inference.engine_options`` sub-block of the YAML body."""
-
-    engine: Optional[inference_engine_module.InferenceEngine] = None
-    engine_args_override: Optional[list[str]] = None
-
-    @field_serializer("engine")
-    def _serialize_engine(self, engine: Optional[inference_engine_module.InferenceEngine]) -> Optional[str]:
-        # The EXECUTE INFERENCE JOB SERVICE engine enum is upper-case
-        # (DEFAULT / VLLM / PYTHON_GENERIC), but the Python enum's value is
-        # lower-case. Emit the member name so the server accepts it.
-        return engine.name if engine is not None else None
-
-
-class Inference(BaseModel):
-    """Inference block of the batch inference YAML body."""
-
-    num_workers: Optional[int] = None
-    max_batch_rows: Optional[int] = None
-    engine_options: Optional[EngineOptions] = None
-
-
-class ImageBuild(BaseModel):
-    """Image-build block of the batch inference YAML body."""
-
-    image_repo: Optional[str] = None
-    force_rebuild: bool = False

{snowflake_ml_python-1.49.0 → snowflake_ml_python-1.50.0}/snowflake/ml/model/_client/model/model_version_impl.py

@@ -14,6 +14,7 @@ from snowflake.ml.jobs import job
 from snowflake.ml.lineage import lineage_node
 from snowflake.ml.model import inference_engine, openai_signatures, task, type_hints
 from snowflake.ml.model._client.model import (
+    batch_inference_job_specs,
     batch_inference_specs,
     inference_engine_utils,
 )
@@ -664,6 +665,12 @@ class ModelVersion(lineage_node.LineageNode):
     ) -> job.MLJob[Any]:
         """Execute batch inference on datasets as an SPCS job.
 
+        .. deprecated::
+            The backend this method uses will be removed in a future release, and it will be
+            replaced by an updated batch inference jobs API. For new work, use
+            :meth:`_run_batch_v2`, which runs on the updated backend and becomes the public
+            ``run_batch`` in the next major release (2.0).
+
         Args:
             compute_pool (str): Name of the compute pool to use for building the image containers and batch
                 inference execution.
@@ -768,6 +775,13 @@ class ModelVersion(lineage_node.LineageNode):
             The input data is temporarily stored in the output stage location under /_temporary before
             inference execution.
         """
+        warnings.warn(
+            "run_batch is deprecated: the batch inference backend it currently uses will be removed in a "
+            "future release, and the method will be replaced by an updated batch inference jobs API. Existing "
+            "calls will stop working once the backend is removed.",
+            DeprecationWarning,
+            stacklevel=2,
+        )
         statement_params = telemetry.get_statement_params(
             project=_TELEMETRY_PROJECT,
             subproject=_TELEMETRY_SUBPROJECT,
@@ -894,26 +908,31 @@ class ModelVersion(lineage_node.LineageNode):
 
     def _run_batch_v2(
         self,
-        X: dataframe.DataFrame,
+        X: Optional[dataframe.DataFrame] = None,
         *,
+        input_stage_location: Optional[str] = None,
         compute_pool: str,
-        output_spec: batch_inference_specs.Output,
-        input_spec: Optional[batch_inference_specs.Input] = None,
-        resources_spec: Optional[batch_inference_specs.Resources] = None,
-        inference_spec: Optional[batch_inference_specs.Inference] = None,
-        image_build_spec: Optional[batch_inference_specs.ImageBuild] = None,
+        output_spec: batch_inference_job_specs.OutputSpec,
+        input_spec: Optional[batch_inference_job_specs.InputSpec] = None,
+        resources_spec: Optional[batch_inference_job_specs.ResourcesSpec] = None,
+        inference_spec: Optional[batch_inference_job_specs.InferenceSpec] = None,
+        image_build_spec: Optional[batch_inference_job_specs.ImageBuildSpec] = None,
         function_name: Optional[str] = None,
         job_name: Optional[str] = None,
         replicas: Optional[int] = None,
         async_: bool = True,
     ) -> job.MLJob[Any]:
-        """Run batch inference via ``EXECUTE INFERENCE JOB SERVICE`` (private).
+        """Run batch inference on a model as a Snowflake job.
 
-        This is a private preview API. Prefer :meth:`run_batch` until this path
-        is generally available.
+        This is the batch inference jobs API that will become the public :meth:`run_batch`
+        in a future release; the deprecated :meth:`run_batch` runs on the older backend and
+        should not be used for new work.
 
         Args:
-            X: Snowpark DataFrame with the input rows.
+            X: Optional Snowpark DataFrame with the input rows. Provide exactly one of ``X`` or
+                ``input_stage_location``.
+            input_stage_location: Optional existing stage path holding the input data. Provide exactly
+                one of ``X`` or ``input_stage_location``.
             compute_pool: Compute pool used by the SPCS job and image build.
             output_spec: Output block. ``stage_location`` is treated as a base;
                 results are written under a per-job subdirectory,
@@ -933,16 +952,19 @@ class ModelVersion(lineage_node.LineageNode):
             MLJob handle for the launched batch inference job.
 
         Raises:
-            ValueError: If ``input_spec.partition_column`` is supplied for a
-                HuggingFace pipeline model or a FUNCTION-type method, or if
-                the partition column collides with a partitioned model output.
+            ValueError: If not exactly one of ``X`` / ``input_stage_location`` is provided, or if
+                ``input_spec.partition_column`` is supplied for a HuggingFace pipeline model or a
+                FUNCTION-type method, or if the partition column collides with a partitioned model output.
         """
         statement_params = telemetry.get_statement_params(
             project=_TELEMETRY_PROJECT,
             subproject=_TELEMETRY_SUBPROJECT,
         )
 
-        effective_input_spec = input_spec if input_spec is not None else batch_inference_specs.Input()
+        if (X is None) == (input_stage_location is None):
+            raise ValueError("Exactly one of X or input_stage_location must be provided.")
+
+        effective_input_spec = input_spec if input_spec is not None else batch_inference_job_specs.InputSpec()
         partition_columns = (
             [effective_input_spec.partition_column] if effective_input_spec.partition_column is not None else None
         )
@@ -984,6 +1006,7 @@ class ModelVersion(lineage_node.LineageNode):
 
         return self._service_ops.execute_inference_job_service(
             X=X,
+            input_stage_location=input_stage_location,
             model_name=self._model_name,
             version_name=self._version_name,
             compute_pool_name=sql_identifier.SqlIdentifier(compute_pool),