[ISSUE] cannot deploy foundation models provisioned throughput endpoint via sdk

[marianreuss]

Description I'm trying to deploy a gpt-oss-120b provisioned throughput endpoint via the sdk (0.69.0). I then run into the following error:

databricks.sdk.errors.platform.InvalidParameterValue: Served entity system.ai.gpt-oss-120b is not eligible for Provisioned Throughput.

However, deploying the foundation model via the UI works just fine.

Reproduction

from databricks.sdk import WorkspaceClient
from databricks.sdk.service.serving import EndpointCoreConfigInput, ServedEntityInput

w = WorkspaceClient()

w.serving_endpoints.create_and_wait(
    name="pt_endpoint",
    config=EndpointCoreConfigInput(
        served_entities=[
            ServedEntityInput(
                name="foo",
                entity_name="system.ai.gpt-oss-120b",
                scale_to_zero_enabled=False,
                entity_version=1,
                min_provisioned_throughput=10,
                max_provisioned_throughput=100,
        )
        ]
    )
)

Expected behavior The model serving endpoint is deployed.

Is it a regression? Not that I know of. Tested some older versions, same result.

Debug Logs

INFO:databricks.sdk:loading DEFAULT profile from ~/.databrickscfg: host, token
DEBUG:databricks.sdk:Attempting to configure auth: pat
DEBUG:urllib3.connectionpool:Starting new HTTPS connection (1): <workspace-url.com>:443
DEBUG:urllib3.connectionpool:<workspace-url.com>:443 "POST /api/2.0/serving-endpoints HTTP/11" 400 None
DEBUG:databricks.sdk:POST /api/2.0/serving-endpoints
> {
>   "config": {
>     "served_entities": [
>       {
>         "entity_name": "system.ai.gpt-oss-120b",
>         "entity_version": 1,
>         "max_provisioned_throughput": 100,
>         "min_provisioned_throughput": 10,
>         "name": "foo",
>         "scale_to_zero_enabled": false
>       }
>     ]
>   },
>   "name": "pt_endpoint"
> }
< 400 Bad Request
< {
<   "details": [
<     {
<       "@type": "type.googleapis.com/google.rpc.RequestInfo",
<       "request_id": "3b1a97ca-27cb-4db7-8975-8e2104617a10",
<       "serving_data": ""
<     }
<   ],
<   "error_code": "INVALID_PARAMETER_VALUE",
<   "message": "Served entity system.ai.gpt-oss-120b is not eligible for Provisioned Throughput."
< }
Traceback (most recent call last):
  File "/Users/user.name/projects/deploy_pt/main.py", line 9, in <module>
    w.serving_endpoints.create_and_wait(
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/service/serving.py", line 4157, in create_and_wait
    return self.create(
           ^^^^^^^^^^^^
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/service/serving.py", line 4136, in create
    op_response = self._api.do("POST", "/api/2.0/serving-endpoints", body=body, headers=headers)
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/core.py", line 85, in do
    return self._api_client.do(
           ^^^^^^^^^^^^^^^^^^^^
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/_base_client.py", line 199, in do
    response = call(
               ^^^^^
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/retries.py", line 59, in wrapper
    raise err
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/retries.py", line 38, in wrapper
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user.name/miniconda3/lib/python3.12/site-packages/databricks/sdk/_base_client.py", line 301, in _perform
    raise error from None
databricks.sdk.errors.platform.InvalidParameterValue: Served entity system.ai.gpt-oss-120b is not eligible for Provisioned Throughput.

Other Information

• OS: macOS
• Version: Sequoia 15.7.1

Additional context

[theDarkDuke]

I'm encountering the same issue