[Bug]: Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported

Start by reading the OpenAI VLM backend path shown in the traceback, especially openviking/models/vlm/backends/openai_vlm.py and its completion request construction. Reproduce with the supplied OpenAI configuration using the reported model and trace where temperature: 0.0 is set. Done when the request no longer fails with the unsupported-temperature error and summary/overview generation succeeds.

Issue Origin

Observed or reproduced in a real environment

Bug Description

2026-10-02 13:00:28,724 - openviking.models.vlm.backends.openai_vlm - WARNING - OpenAI VLM async completion failed with retryable error (retry 2/3): Connection error.; retrying in 1.22s
2026-10-02 13:00:28,796 - openviking.storage.queuefs.semantic_dag - WARNING - Failed to generate summary for viking://resources/amp-glcs/AMP_GLCS_Complete_Documentation.md: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}
2026-10-02 13:00:30,587 - openviking.storage.queuefs.semantic_dag - WARNING - Failed to generate summary for viking://resources/amp-glcs/PGEIN-8869_ClimateControl_ETS_EpochTimestamp.md: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}
2026-10-02 13:00:31,718 - openviking.storage.queuefs.semantic_processor - ERROR - Failed to generate overview for viking://resources/amp-glcs: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}
Traceback (most recent call last):
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\storage\queuefs\semantic_processor.py", line 1422, in _single_generate_overview
overview = await vlm.get_completion_async(prompt)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking_cli\utils\config\vlm_config.py", line 723, in get_completion_async
return await self.get_vlm_instance().get_completion_async(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\models\vlm\backends\openai_vlm.py", line 368, in get_completion_async
return await retry_async(
^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\utils\model_retry.py", line 442, in retry_async
return await func()
^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\models\vlm\backends\openai_vlm.py", line 355, in _call
response = await client.chat.completions.create(**kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai\resources\chat\completions\completions.py", line 2907, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai_base_client.py", line 1992, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai_base_client.py", line 1777, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}
2026-10-02 13:00:32,750 - openviking.storage.queuefs.semantic_processor - ERROR - Failed to generate overview for viking://resources: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}
Traceback (most recent call last):
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\storage\queuefs\semantic_processor.py", line 1422, in _single_generate_overview
overview = await vlm.get_completion_async(prompt)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking_cli\utils\config\vlm_config.py", line 723, in get_completion_async
return await self.get_vlm_instance().get_completion_async(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\models\vlm\backends\openai_vlm.py", line 368, in get_completion_async
return await retry_async(
^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\utils\model_retry.py", line 442, in retry_async
return await func()
^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openviking\models\vlm\backends\openai_vlm.py", line 355, in _call
response = await client.chat.completions.create(**kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai\resources\chat\completions\completions.py", line 2907, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai_base_client.py", line 1992, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\hflbt\AppData\Roaming\Python\Python312\site-packages\openai_base_client.py", line 1777, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': "Unsupported value: 'temperature' does not support 0.0 with this model. Only the default (1) value is supported.", 'type': 'invalid_request_error', 'param': 'temperature', 'code': 'unsupported_value'}}

Steps to Reproduce

Ask copilot agent to write/read data to openviking
{
"embedding": {
"dense": {
"provider": "openai",
"model": "text-embedding-3-small",
"api_key": "sk-proj-....",
"dimension": 1536
}
},
"vlm": {
"provider": "openai",
"model": "gpt-6-sol",
"api_key": "sk-proj-...."
}
}

Expected Behavior

No error

Actual Behavior

Some error in logs

Minimal Reproducible Example

Error Logs

OpenViking Version

latest

Python Version

3.12

Operating System

Windows

Model Backend

OpenAI

Additional Context

No WSL