elasticsearch 9.2.0__py3-none-any.whl → 9.3.0__py3-none-any.whl

elasticsearch/_async/client/inference.py

@@ -44,14 +44,20 @@ class InferenceClient(NamespacedClient):
         """
         .. raw:: html
 
-          <p>Perform completion inference on the service</p>
+          <p>Perform completion inference on the service.</p>
+          <p>Get responses for completion tasks.
+          This API works only with the completion task type.</p>
+          <p>IMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Azure, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face. For built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.</p>
+          <p>This API requires the <code>monitor_inference</code> cluster privilege (the built-in <code>inference_admin</code> and <code>inference_user</code> roles grant this privilege).</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-inference>`_
 
         :param inference_id: The inference Id
         :param input: Inference input. Either a string or an array of strings.
-        :param task_settings: Optional task settings
+        :param task_settings: Task settings for the individual inference request. These
+            settings are specific to the <task_type> you specified and override the task
+            settings specified when initializing the service.
         :param timeout: Specifies the amount of time to wait for the inference request
             to complete.
         """
@@ -116,15 +122,17 @@ class InferenceClient(NamespacedClient):
         """
         .. raw:: html
 
-          <p>Delete an inference endpoint</p>
+          <p>Delete an inference endpoint.</p>
+          <p>This API requires the manage_inference cluster privilege (the built-in <code>inference_admin</code> role grants this privilege).</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-delete>`_
 
         :param inference_id: The inference identifier.
         :param task_type: The task type
-        :param dry_run: When true, the endpoint is not deleted and a list of ingest processors
-            which reference this endpoint is returned.
+        :param dry_run: When true, checks the semantic_text fields and inference processors
+            that reference the endpoint and returns them in a list, but does not delete
+            the endpoint.
         :param force: When true, the inference endpoint is forcefully deleted even if
             it is still being used by ingest processors or semantic text fields.
         """
@@ -190,7 +198,8 @@ class InferenceClient(NamespacedClient):
         """
         .. raw:: html
 
-          <p>Get an inference endpoint</p>
+          <p>Get an inference endpoint.</p>
+          <p>This API requires the <code>monitor_inference</code> cluster privilege (the built-in <code>inference_admin</code> and <code>inference_user</code> roles grant this privilege).</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-get>`_
@@ -386,19 +395,22 @@ class InferenceClient(NamespacedClient):
           <li>Amazon Bedrock (<code>completion</code>, <code>text_embedding</code>)</li>
           <li>Amazon SageMaker (<code>chat_completion</code>, <code>completion</code>, <code>rerank</code>, <code>sparse_embedding</code>, <code>text_embedding</code>)</li>
           <li>Anthropic (<code>completion</code>)</li>
-          <li>Azure AI Studio (<code>completion</code>, 'rerank', <code>text_embedding</code>)</li>
-          <li>Azure OpenAI (<code>completion</code>, <code>text_embedding</code>)</li>
+          <li>Azure AI Studio (<code>completion</code>, <code>rerank</code>, <code>text_embedding</code>)</li>
+          <li>Azure OpenAI (<code>chat_completion</code>, <code>completion</code>, <code>text_embedding</code>)</li>
           <li>Cohere (<code>completion</code>, <code>rerank</code>, <code>text_embedding</code>)</li>
           <li>DeepSeek (<code>chat_completion</code>, <code>completion</code>)</li>
           <li>Elasticsearch (<code>rerank</code>, <code>sparse_embedding</code>, <code>text_embedding</code> - this service is for built-in models and models uploaded through Eland)</li>
           <li>ELSER (<code>sparse_embedding</code>)</li>
           <li>Google AI Studio (<code>completion</code>, <code>text_embedding</code>)</li>
           <li>Google Vertex AI (<code>chat_completion</code>, <code>completion</code>, <code>rerank</code>, <code>text_embedding</code>)</li>
+          <li>Groq (<code>chat_completion</code>)</li>
           <li>Hugging Face (<code>chat_completion</code>, <code>completion</code>, <code>rerank</code>, <code>text_embedding</code>)</li>
           <li>JinaAI (<code>rerank</code>, <code>text_embedding</code>)</li>
           <li>Llama (<code>chat_completion</code>, <code>completion</code>, <code>text_embedding</code>)</li>
           <li>Mistral (<code>chat_completion</code>, <code>completion</code>, <code>text_embedding</code>)</li>
+          <li>Nvidia (<code>chat_completion</code>, <code>completion</code>, <code>text_embedding</code>, <code>rerank</code>)</li>
           <li>OpenAI (<code>chat_completion</code>, <code>completion</code>, <code>text_embedding</code>)</li>
+          <li>OpenShift AI (<code>chat_completion</code>, <code>completion</code>, <code>rerank</code>, <code>text_embedding</code>)</li>
           <li>VoyageAI (<code>rerank</code>, <code>text_embedding</code>)</li>
           <li>Watsonx inference integration (<code>text_embedding</code>)</li>
           </ul>
@@ -544,7 +556,7 @@ class InferenceClient(NamespacedClient):
         self,
         *,
         task_type: t.Union[
-            str, t.Literal["completion", "rerank", "space_embedding", "text_embedding"]
+            str, t.Literal["completion", "rerank", "sparse_embedding", "text_embedding"]
         ],
         alibabacloud_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["alibabacloud-ai-search"]]] = None,
@@ -573,7 +585,9 @@ class InferenceClient(NamespacedClient):
             this case, `alibabacloud-ai-search`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `alibabacloud-ai-search` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `sparse_embedding` or `text_embedding` task types. Not applicable to
+            the `rerank` or `completion` task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -669,7 +683,8 @@ class InferenceClient(NamespacedClient):
             this case, `amazonbedrock`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `amazonbedrock` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` task type.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -771,7 +786,9 @@ class InferenceClient(NamespacedClient):
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `amazon_sagemaker` service and `service_settings.api`
             you specified.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `sparse_embedding` or `text_embedding` task types. Not applicable to
+            the `rerank`, `completion`, or `chat_completion` task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type and `service_settings.api` you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -825,12 +842,7 @@ class InferenceClient(NamespacedClient):
         )
 
     @_rewrite_parameters(
-        body_fields=(
-            "service",
-            "service_settings",
-            "chunking_settings",
-            "task_settings",
-        ),
+        body_fields=("service", "service_settings", "task_settings"),
     )
     async def put_anthropic(
         self,
@@ -839,7 +851,6 @@ class InferenceClient(NamespacedClient):
         anthropic_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["anthropic"]]] = None,
         service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
-        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
         error_trace: t.Optional[bool] = None,
         filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
         human: t.Optional[bool] = None,
@@ -863,8 +874,7 @@ class InferenceClient(NamespacedClient):
         :param service: The type of service supported for the specified task type. In
             this case, `anthropic`.
         :param service_settings: Settings used to install the inference model. These
-            settings are specific to the `watsonxai` service.
-        :param chunking_settings: The chunking configuration object.
+            settings are specific to the `anthropic` service.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -902,8 +912,6 @@ class InferenceClient(NamespacedClient):
                 __body["service"] = service
             if service_settings is not None:
                 __body["service_settings"] = service_settings
-            if chunking_settings is not None:
-                __body["chunking_settings"] = chunking_settings
             if task_settings is not None:
                 __body["task_settings"] = task_settings
         __headers = {"accept": "application/json", "content-type": "application/json"}
@@ -955,8 +963,10 @@ class InferenceClient(NamespacedClient):
         :param service: The type of service supported for the specified task type. In
             this case, `azureaistudio`.
         :param service_settings: Settings used to install the inference model. These
-            settings are specific to the `openai` service.
-        :param chunking_settings: The chunking configuration object.
+            settings are specific to the `azureaistudio` service.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank` or `completion`
+            task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1020,7 +1030,9 @@ class InferenceClient(NamespacedClient):
     async def put_azureopenai(
         self,
         *,
-        task_type: t.Union[str, t.Literal["completion", "text_embedding"]],
+        task_type: t.Union[
+            str, t.Literal["chat_completion", "completion", "text_embedding"]
+        ],
         azureopenai_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["azureopenai"]]] = None,
         service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
@@ -1056,7 +1068,9 @@ class InferenceClient(NamespacedClient):
             this case, `azureopenai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `azureopenai` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` and `chat_completion`
+            task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1148,7 +1162,9 @@ class InferenceClient(NamespacedClient):
             this case, `cohere`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `cohere` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank` or `completion`
+            task type.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1200,12 +1216,7 @@ class InferenceClient(NamespacedClient):
         )
 
     @_rewrite_parameters(
-        body_fields=(
-            "service",
-            "service_settings",
-            "chunking_settings",
-            "task_settings",
-        ),
+        body_fields=("service", "service_settings", "task_settings"),
     )
     async def put_contextualai(
         self,
@@ -1214,7 +1225,6 @@ class InferenceClient(NamespacedClient):
         contextualai_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["contextualai"]]] = None,
         service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
-        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
         error_trace: t.Optional[bool] = None,
         filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
         human: t.Optional[bool] = None,
@@ -1239,7 +1249,6 @@ class InferenceClient(NamespacedClient):
             this case, `contextualai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `contextualai` service.
-        :param chunking_settings: The chunking configuration object.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1277,8 +1286,6 @@ class InferenceClient(NamespacedClient):
                 __body["service"] = service
             if service_settings is not None:
                 __body["service_settings"] = service_settings
-            if chunking_settings is not None:
-                __body["chunking_settings"] = chunking_settings
             if task_settings is not None:
                 __body["task_settings"] = task_settings
         __headers = {"accept": "application/json", "content-type": "application/json"}
@@ -1372,7 +1379,9 @@ class InferenceClient(NamespacedClient):
             this case, `custom`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `custom` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `sparse_embedding` or `text_embedding` task types. Not applicable to
+            the `rerank` or `completion` task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         """
@@ -1420,7 +1429,7 @@ class InferenceClient(NamespacedClient):
         )
 
     @_rewrite_parameters(
-        body_fields=("service", "service_settings", "chunking_settings"),
+        body_fields=("service", "service_settings"),
     )
     async def put_deepseek(
         self,
@@ -1429,7 +1438,6 @@ class InferenceClient(NamespacedClient):
         deepseek_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["deepseek"]]] = None,
         service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
-        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
         error_trace: t.Optional[bool] = None,
         filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
         human: t.Optional[bool] = None,
@@ -1452,7 +1460,6 @@ class InferenceClient(NamespacedClient):
             this case, `deepseek`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `deepseek` service.
-        :param chunking_settings: The chunking configuration object.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
             to be created.
         """
@@ -1486,8 +1493,6 @@ class InferenceClient(NamespacedClient):
                 __body["service"] = service
             if service_settings is not None:
                 __body["service_settings"] = service_settings
-            if chunking_settings is not None:
-                __body["chunking_settings"] = chunking_settings
         __headers = {"accept": "application/json", "content-type": "application/json"}
         return await self.perform_request(  # type: ignore[return-value]
             "PUT",
@@ -1554,7 +1559,9 @@ class InferenceClient(NamespacedClient):
             this case, `elasticsearch`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `elasticsearch` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `sparse_embedding` and `text_embedding` task types. Not applicable to
+            the `rerank` task type.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1735,7 +1742,8 @@ class InferenceClient(NamespacedClient):
             this case, `googleaistudio`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `googleaistudio` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` task type.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
             to be created.
         """
@@ -1825,7 +1833,9 @@ class InferenceClient(NamespacedClient):
             this case, `googlevertexai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `googlevertexai` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank`, `completion`,
+            or `chat_completion` task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -1878,6 +1888,82 @@ class InferenceClient(NamespacedClient):
             path_parts=__path_parts,
         )
 
+    @_rewrite_parameters(
+        body_fields=("service", "service_settings"),
+    )
+    async def put_groq(
+        self,
+        *,
+        task_type: t.Union[str, t.Literal["chat_completion"]],
+        groq_inference_id: str,
+        service: t.Optional[t.Union[str, t.Literal["groq"]]] = None,
+        service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        error_trace: t.Optional[bool] = None,
+        filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
+        human: t.Optional[bool] = None,
+        pretty: t.Optional[bool] = None,
+        timeout: t.Optional[t.Union[str, t.Literal[-1], t.Literal[0]]] = None,
+        body: t.Optional[t.Dict[str, t.Any]] = None,
+    ) -> ObjectApiResponse[t.Any]:
+        """
+        .. raw:: html
+
+          <p>Create a Groq inference endpoint.</p>
+          <p>Create an inference endpoint to perform an inference task with the <code>groq</code> service.</p>
+
+
+        `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-put-groq>`_
+
+        :param task_type: The type of the inference task that the model will perform.
+        :param groq_inference_id: The unique identifier of the inference endpoint.
+        :param service: The type of service supported for the specified task type. In
+            this case, `groq`.
+        :param service_settings: Settings used to install the inference model. These
+            settings are specific to the `groq` service.
+        :param timeout: Specifies the amount of time to wait for the inference endpoint
+            to be created.
+        """
+        if task_type in SKIP_IN_PATH:
+            raise ValueError("Empty value passed for parameter 'task_type'")
+        if groq_inference_id in SKIP_IN_PATH:
+            raise ValueError("Empty value passed for parameter 'groq_inference_id'")
+        if service is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service'")
+        if service_settings is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service_settings'")
+        __path_parts: t.Dict[str, str] = {
+            "task_type": _quote(task_type),
+            "groq_inference_id": _quote(groq_inference_id),
+        }
+        __path = f'/_inference/{__path_parts["task_type"]}/{__path_parts["groq_inference_id"]}'
+        __query: t.Dict[str, t.Any] = {}
+        __body: t.Dict[str, t.Any] = body if body is not None else {}
+        if error_trace is not None:
+            __query["error_trace"] = error_trace
+        if filter_path is not None:
+            __query["filter_path"] = filter_path
+        if human is not None:
+            __query["human"] = human
+        if pretty is not None:
+            __query["pretty"] = pretty
+        if timeout is not None:
+            __query["timeout"] = timeout
+        if not __body:
+            if service is not None:
+                __body["service"] = service
+            if service_settings is not None:
+                __body["service_settings"] = service_settings
+        __headers = {"accept": "application/json", "content-type": "application/json"}
+        return await self.perform_request(  # type: ignore[return-value]
+            "PUT",
+            __path,
+            params=__query,
+            headers=__headers,
+            body=__body,
+            endpoint_id="inference.put_groq",
+            path_parts=__path_parts,
+        )
+
     @_rewrite_parameters(
         body_fields=(
             "service",
@@ -1953,7 +2039,9 @@ class InferenceClient(NamespacedClient):
             this case, `hugging_face`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `hugging_face` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank`, `completion`,
+            or `chat_completion` task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -2047,7 +2135,8 @@ class InferenceClient(NamespacedClient):
             this case, `jinaai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `jinaai` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank` task type.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -2133,7 +2222,9 @@ class InferenceClient(NamespacedClient):
             this case, `llama`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `llama` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` or `chat_completion`
+            task types.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
             to be created.
         """
@@ -2215,7 +2306,9 @@ class InferenceClient(NamespacedClient):
             this case, `mistral`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `mistral` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` or `chat_completion`
+            task types.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
             to be created.
         """
@@ -2262,6 +2355,104 @@ class InferenceClient(NamespacedClient):
             path_parts=__path_parts,
         )
 
+    @_rewrite_parameters(
+        body_fields=(
+            "service",
+            "service_settings",
+            "chunking_settings",
+            "task_settings",
+        ),
+    )
+    async def put_nvidia(
+        self,
+        *,
+        task_type: t.Union[
+            str, t.Literal["chat_completion", "completion", "rerank", "text_embedding"]
+        ],
+        nvidia_inference_id: str,
+        service: t.Optional[t.Union[str, t.Literal["nvidia"]]] = None,
+        service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        error_trace: t.Optional[bool] = None,
+        filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
+        human: t.Optional[bool] = None,
+        pretty: t.Optional[bool] = None,
+        task_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        timeout: t.Optional[t.Union[str, t.Literal[-1], t.Literal[0]]] = None,
+        body: t.Optional[t.Dict[str, t.Any]] = None,
+    ) -> ObjectApiResponse[t.Any]:
+        """
+        .. raw:: html
+
+          <p>Create an Nvidia inference endpoint.</p>
+          <p>Create an inference endpoint to perform an inference task with the <code>nvidia</code> service.</p>
+
+
+        `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-put-nvidia>`_
+
+        :param task_type: The type of the inference task that the model will perform.
+            NOTE: The `chat_completion` task type only supports streaming and only through
+            the _stream API.
+        :param nvidia_inference_id: The unique identifier of the inference endpoint.
+        :param service: The type of service supported for the specified task type. In
+            this case, `nvidia`.
+        :param service_settings: Settings used to install the inference model. These
+            settings are specific to the `nvidia` service.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank`, `completion`,
+            or `chat_completion` task types.
+        :param task_settings: Settings to configure the inference task. Applies only
+            to the `text_embedding` task type. Not applicable to the `rerank`, `completion`,
+            or `chat_completion` task types. These settings are specific to the task
+            type you specified.
+        :param timeout: Specifies the amount of time to wait for the inference endpoint
+            to be created.
+        """
+        if task_type in SKIP_IN_PATH:
+            raise ValueError("Empty value passed for parameter 'task_type'")
+        if nvidia_inference_id in SKIP_IN_PATH:
+            raise ValueError("Empty value passed for parameter 'nvidia_inference_id'")
+        if service is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service'")
+        if service_settings is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service_settings'")
+        __path_parts: t.Dict[str, str] = {
+            "task_type": _quote(task_type),
+            "nvidia_inference_id": _quote(nvidia_inference_id),
+        }
+        __path = f'/_inference/{__path_parts["task_type"]}/{__path_parts["nvidia_inference_id"]}'
+        __query: t.Dict[str, t.Any] = {}
+        __body: t.Dict[str, t.Any] = body if body is not None else {}
+        if error_trace is not None:
+            __query["error_trace"] = error_trace
+        if filter_path is not None:
+            __query["filter_path"] = filter_path
+        if human is not None:
+            __query["human"] = human
+        if pretty is not None:
+            __query["pretty"] = pretty
+        if timeout is not None:
+            __query["timeout"] = timeout
+        if not __body:
+            if service is not None:
+                __body["service"] = service
+            if service_settings is not None:
+                __body["service_settings"] = service_settings
+            if chunking_settings is not None:
+                __body["chunking_settings"] = chunking_settings
+            if task_settings is not None:
+                __body["task_settings"] = task_settings
+        __headers = {"accept": "application/json", "content-type": "application/json"}
+        return await self.perform_request(  # type: ignore[return-value]
+            "PUT",
+            __path,
+            params=__query,
+            headers=__headers,
+            body=__body,
+            endpoint_id="inference.put_nvidia",
+            path_parts=__path_parts,
+        )
+
     @_rewrite_parameters(
         body_fields=(
             "service",
@@ -2305,7 +2496,9 @@ class InferenceClient(NamespacedClient):
             this case, `openai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `openai` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` or `chat_completion`
+            task types.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -2356,6 +2549,106 @@ class InferenceClient(NamespacedClient):
             path_parts=__path_parts,
         )
 
+    @_rewrite_parameters(
+        body_fields=(
+            "service",
+            "service_settings",
+            "chunking_settings",
+            "task_settings",
+        ),
+    )
+    async def put_openshift_ai(
+        self,
+        *,
+        task_type: t.Union[
+            str, t.Literal["chat_completion", "completion", "rerank", "text_embedding"]
+        ],
+        openshiftai_inference_id: str,
+        service: t.Optional[t.Union[str, t.Literal["openshift_ai"]]] = None,
+        service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        error_trace: t.Optional[bool] = None,
+        filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
+        human: t.Optional[bool] = None,
+        pretty: t.Optional[bool] = None,
+        task_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        timeout: t.Optional[t.Union[str, t.Literal[-1], t.Literal[0]]] = None,
+        body: t.Optional[t.Dict[str, t.Any]] = None,
+    ) -> ObjectApiResponse[t.Any]:
+        """
+        .. raw:: html
+
+          <p>Create an OpenShift AI inference endpoint.</p>
+          <p>Create an inference endpoint to perform an inference task with the <code>openshift_ai</code> service.</p>
+
+
+        `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-put-openshift-ai>`_
+
+        :param task_type: The type of the inference task that the model will perform.
+            NOTE: The `chat_completion` task type only supports streaming and only through
+            the _stream API.
+        :param openshiftai_inference_id: The unique identifier of the inference endpoint.
+        :param service: The type of service supported for the specified task type. In
+            this case, `openshift_ai`.
+        :param service_settings: Settings used to install the inference model. These
+            settings are specific to the `openshift_ai` service.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank`, `completion`,
+            or `chat_completion` task types.
+        :param task_settings: Settings to configure the inference task. Applies only
+            to the `rerank` task type. Not applicable to the `text_embedding`, `completion`,
+            or `chat_completion` task types. These settings are specific to the task
+            type you specified.
+        :param timeout: Specifies the amount of time to wait for the inference endpoint
+            to be created.
+        """
+        if task_type in SKIP_IN_PATH:
+            raise ValueError("Empty value passed for parameter 'task_type'")
+        if openshiftai_inference_id in SKIP_IN_PATH:
+            raise ValueError(
+                "Empty value passed for parameter 'openshiftai_inference_id'"
+            )
+        if service is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service'")
+        if service_settings is None and body is None:
+            raise ValueError("Empty value passed for parameter 'service_settings'")
+        __path_parts: t.Dict[str, str] = {
+            "task_type": _quote(task_type),
+            "openshiftai_inference_id": _quote(openshiftai_inference_id),
+        }
+        __path = f'/_inference/{__path_parts["task_type"]}/{__path_parts["openshiftai_inference_id"]}'
+        __query: t.Dict[str, t.Any] = {}
+        __body: t.Dict[str, t.Any] = body if body is not None else {}
+        if error_trace is not None:
+            __query["error_trace"] = error_trace
+        if filter_path is not None:
+            __query["filter_path"] = filter_path
+        if human is not None:
+            __query["human"] = human
+        if pretty is not None:
+            __query["pretty"] = pretty
+        if timeout is not None:
+            __query["timeout"] = timeout
+        if not __body:
+            if service is not None:
+                __body["service"] = service
+            if service_settings is not None:
+                __body["service_settings"] = service_settings
+            if chunking_settings is not None:
+                __body["chunking_settings"] = chunking_settings
+            if task_settings is not None:
+                __body["task_settings"] = task_settings
+        __headers = {"accept": "application/json", "content-type": "application/json"}
+        return await self.perform_request(  # type: ignore[return-value]
+            "PUT",
+            __path,
+            params=__query,
+            headers=__headers,
+            body=__body,
+            endpoint_id="inference.put_openshift_ai",
+            path_parts=__path_parts,
+        )
+
     @_rewrite_parameters(
         body_fields=(
             "service",
@@ -2396,7 +2689,8 @@ class InferenceClient(NamespacedClient):
             this case, `voyageai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `voyageai` service.
-        :param chunking_settings: The chunking configuration object.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `rerank` task type.
         :param task_settings: Settings to configure the inference task. These settings
             are specific to the task type you specified.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
@@ -2448,7 +2742,7 @@ class InferenceClient(NamespacedClient):
         )
 
     @_rewrite_parameters(
-        body_fields=("service", "service_settings"),
+        body_fields=("service", "service_settings", "chunking_settings"),
     )
     async def put_watsonx(
         self,
@@ -2459,6 +2753,7 @@ class InferenceClient(NamespacedClient):
         watsonx_inference_id: str,
         service: t.Optional[t.Union[str, t.Literal["watsonxai"]]] = None,
         service_settings: t.Optional[t.Mapping[str, t.Any]] = None,
+        chunking_settings: t.Optional[t.Mapping[str, t.Any]] = None,
         error_trace: t.Optional[bool] = None,
         filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
         human: t.Optional[bool] = None,
@@ -2483,6 +2778,9 @@ class InferenceClient(NamespacedClient):
             this case, `watsonxai`.
         :param service_settings: Settings used to install the inference model. These
             settings are specific to the `watsonxai` service.
+        :param chunking_settings: The chunking configuration object. Applies only to
+            the `text_embedding` task type. Not applicable to the `completion` or `chat_completion`
+            task types.
         :param timeout: Specifies the amount of time to wait for the inference endpoint
             to be created.
         """
@@ -2516,6 +2814,8 @@ class InferenceClient(NamespacedClient):
                 __body["service"] = service
             if service_settings is not None:
                 __body["service_settings"] = service_settings
+            if chunking_settings is not None:
+                __body["chunking_settings"] = chunking_settings
         __headers = {"accept": "application/json", "content-type": "application/json"}
         return await self.perform_request(  # type: ignore[return-value]
             "PUT",
@@ -2528,39 +2828,41 @@ class InferenceClient(NamespacedClient):
         )
 
     @_rewrite_parameters(
-        body_fields=("input", "query", "task_settings"),
+        body_fields=("input", "query", "return_documents", "task_settings", "top_n"),
     )
     async def rerank(
         self,
         *,
         inference_id: str,
-        input: t.Optional[t.Union[str, t.Sequence[str]]] = None,
+        input: t.Optional[t.Sequence[str]] = None,
         query: t.Optional[str] = None,
         error_trace: t.Optional[bool] = None,
         filter_path: t.Optional[t.Union[str, t.Sequence[str]]] = None,
         human: t.Optional[bool] = None,
         pretty: t.Optional[bool] = None,
+        return_documents: t.Optional[bool] = None,
         task_settings: t.Optional[t.Any] = None,
         timeout: t.Optional[t.Union[str, t.Literal[-1], t.Literal[0]]] = None,
+        top_n: t.Optional[int] = None,
         body: t.Optional[t.Dict[str, t.Any]] = None,
     ) -> ObjectApiResponse[t.Any]:
         """
         .. raw:: html
 
-          <p>Perform reranking inference on the service</p>
+          <p>Perform reranking inference on the service.</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-inference>`_
 
         :param inference_id: The unique identifier for the inference endpoint.
-        :param input: The text on which you want to perform the inference task. It can
-            be a single string or an array. > info > Inference endpoints for the `completion`
-            task type currently only support a single string as input.
+        :param input: The documents to rank.
         :param query: Query input.
+        :param return_documents: Include the document text in the response.
         :param task_settings: Task settings for the individual inference request. These
             settings are specific to the task type you specified and override the task
             settings specified when initializing the service.
         :param timeout: The amount of time to wait for the inference request to complete.
+        :param top_n: Limit the response to the top N documents.
         """
         if inference_id in SKIP_IN_PATH:
             raise ValueError("Empty value passed for parameter 'inference_id'")
@@ -2587,8 +2889,12 @@ class InferenceClient(NamespacedClient):
                 __body["input"] = input
             if query is not None:
                 __body["query"] = query
+            if return_documents is not None:
+                __body["return_documents"] = return_documents
             if task_settings is not None:
                 __body["task_settings"] = task_settings
+            if top_n is not None:
+                __body["top_n"] = top_n
         __headers = {"accept": "application/json", "content-type": "application/json"}
         return await self.perform_request(  # type: ignore[return-value]
             "POST",
@@ -2619,14 +2925,16 @@ class InferenceClient(NamespacedClient):
         """
         .. raw:: html
 
-          <p>Perform sparse embedding inference on the service</p>
+          <p>Perform sparse embedding inference on the service.</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-inference>`_
 
         :param inference_id: The inference Id
         :param input: Inference input. Either a string or an array of strings.
-        :param task_settings: Optional task settings
+        :param task_settings: Task settings for the individual inference request. These
+            settings are specific to the <task_type> you specified and override the task
+            settings specified when initializing the service.
         :param timeout: Specifies the amount of time to wait for the inference request
             to complete.
         """
@@ -2684,7 +2992,7 @@ class InferenceClient(NamespacedClient):
         """
         .. raw:: html
 
-          <p>Perform text embedding inference on the service</p>
+          <p>Perform text embedding inference on the service.</p>
 
 
         `<https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-inference-inference>`_
@@ -2698,7 +3006,9 @@ class InferenceClient(NamespacedClient):
             to the relevant service-specific documentation for more info. > info > The
             `input_type` parameter specified on the root level of the request body will
             take precedence over the `input_type` parameter specified in `task_settings`.
-        :param task_settings: Optional task settings
+        :param task_settings: Task settings for the individual inference request. These
+            settings are specific to the <task_type> you specified and override the task
+            settings specified when initializing the service.
         :param timeout: Specifies the amount of time to wait for the inference request
             to complete.
         """