diff --git a/centml/sdk/api.py b/centml/sdk/api.py
index 3419652..d7c4457 100644
--- a/centml/sdk/api.py
+++ b/centml/sdk/api.py
@@ -12,7 +12,6 @@
     DeploymentStatus,
     CreateInferenceV3DeploymentRequest,
     CreateComputeDeploymentRequest,
-    CreateCServeV3DeploymentRequest,
     CreateDynamoDeploymentRequest,
     CreateJobDeploymentRequest,
     CreateHardwareInstanceRequest,
@@ -169,9 +168,6 @@ def create_compute(self, request: CreateComputeDeploymentRequest):
     def create_job(self, request: CreateJobDeploymentRequest):
         return self._api.create_job_deployment_deployments_job_post(request)
 
-    def create_cserve(self, request: CreateCServeV3DeploymentRequest):
-        return self._api.create_cserve_v3_deployment_deployments_cserve_v3_post(request)
-
     def create_dynamo(self, request: CreateDynamoDeploymentRequest):
         return self._api.create_dynamo_deployment_deployments_dynamo_post(request)
 
@@ -181,9 +177,6 @@ def update_inference(self, deployment_id: int, request: CreateInferenceV3Deploym
     def update_compute(self, deployment_id: int, request: CreateComputeDeploymentRequest):
         return self._api.update_compute_deployment_deployments_compute_put(deployment_id, request)
 
-    def update_cserve(self, deployment_id: int, request: CreateCServeV3DeploymentRequest):
-        return self._api.update_cserve_v3_deployment_deployments_cserve_v3_put(deployment_id, request)
-
     def update_dynamo(self, deployment_id: int, request: CreateDynamoDeploymentRequest):
         return self._api.update_dynamo_deployment_deployments_dynamo_put(deployment_id, request)
 
@@ -220,9 +213,6 @@ def delete_hardware_instance(self, hardware_instance_id: int):
     def get_prebuilt_images(self, depl_type: DeploymentType):
         return self._api.get_prebuilt_images_prebuilt_images_get(type=depl_type)
 
-    def get_cserve_recipe(self, model=None, hf_token=None):
-        return self._api.get_cserve_recipe_deployments_cserve_recipes_get(model=model, hf_token=hf_token).results
-
     def get_cluster_id(self, hardware_instance_id):
         filtered_hw = list(filter(lambda h: h.id == hardware_instance_id, self.get_hardware_instances()))
 
diff --git a/examples/sdk/create_cserve.py b/examples/sdk/create_cserve.py
deleted file mode 100644
index d355019..0000000
--- a/examples/sdk/create_cserve.py
+++ /dev/null
@@ -1,73 +0,0 @@
-import centml
-from centml.sdk.api import get_centml_client
-from centml.sdk import DeploymentType, CreateCServeV3DeploymentRequest, CServeV2Recipe
-
-
-def get_fastest_cserve_config(cclient, name, model):
-    fastest = cclient.get_cserve_recipe(model=model)[0].fastest
-
-    return CreateCServeV3DeploymentRequest(
-        name=name,
-        cluster_id=cclient.get_cluster_id(fastest.hardware_instance_id),
-        hardware_instance_id=fastest.hardware_instance_id,
-        recipe=fastest.recipe,
-        min_replicas=1,
-        max_replicas=1,
-        env_vars={},
-    )
-
-
-def get_default_cserve_config(cclient, name, model):
-    default_recipe = CServeV2Recipe(model=model)
-
-    hardware_instance = cclient.get_hardware_instances(cluster_id=1001)[0]
-
-    return CreateCServeV3DeploymentRequest(
-        name=name,
-        cluster_id=hardware_instance.cluster_id,
-        hardware_instance_id=hardware_instance.id,
-        recipe=default_recipe,
-        min_replicas=1,
-        max_replicas=1,
-        env_vars={},
-    )
-
-
-def main():
-    with get_centml_client() as cclient:
-        ### Get the configurations for the Qwen model
-        qwen_config = get_fastest_cserve_config(cclient, name="qwen-fastest", model="Qwen/Qwen2-VL-7B-Instruct")
-        # qwen_config = get_default_cserve_config(cclient, name="qwen-default", model="Qwen/Qwen2-VL-7B-Instruct")
-
-        ### Modify the recipe if necessary
-        qwen_config.recipe.additional_properties["max_num_seqs"] = 512
-
-        # Create CServeV3 deployment
-        response = cclient.create_cserve(qwen_config)
-        print("Create deployment response: ", response)
-
-        ### Get deployment details
-        deployment = cclient.get_cserve(response.id)  # Automatically detects V2
-        print("Deployment details: ", deployment)
-
-        """
-        ### Scale replicas in place (no new revision / no rolling update)
-        # Changing only min_replicas and/or max_replicas on the same Create*
-        # request updates the selected revision in place; revision_number stays put.
-        # Any other field change still creates a new revision.
-        qwen_config.min_replicas = 2
-        qwen_config.max_replicas = 4
-        cclient.update_cserve(deployment.id, qwen_config)
-        scaled = cclient.get_cserve(deployment.id)
-        print(f"Scaled to {scaled.min_replicas}-{scaled.max_replicas}; revision {scaled.revision_number}")
-
-        ### Pause the deployment
-        cclient.pause(deployment.id)
-
-        ### Delete the deployment
-        cclient.delete(deployment.id)
-        """
-
-
-if __name__ == "__main__":
-    main()
diff --git a/tests/test_cli_cluster.py b/tests/test_cli_cluster.py
index e4cb501..5e3b00e 100644
--- a/tests/test_cli_cluster.py
+++ b/tests/test_cli_cluster.py
@@ -158,7 +158,7 @@ def test_get_job_routes_to_job_api_and_displays_job_config():
 
     deployment = _deployment(status=DeploymentStatus.ACTIVE)
     hardware = HardwareInstanceResponse(
-        id=2, name="h100", gpu_type="H100", num_gpu=8, cpu=96, memory=1024, cluster_id=1
+        id=2, name="h100", gpu_type="H100", num_gpu=8, cpu=96, memory=1024, cost_per_hr=0, cluster_id=1
     )
     runner = CliRunner()