[Bug]: Random Invalid Model Name Errors

[decent-engineer-decent-datascientist]

What happened?

I keep getting this error intermittently when trying to interact with any Azure OpenAI Endpoint that has thinking built in:

Error fetching response:Error: 400 {'error': '/chat/completions: Invalid model name passed in model=None. Call /v1/models to view available models for your key.'}

This is occuring via curl, in LiteLLM's "Test Key" UI, and in my connected openwebui instance. I've confirmed that my Azure Endpoint, called directly does not show these errors.

I've tested this on the following releases:

• v1.75.5.dev3
• litellm_v1.75.5-dev_memory_fix_1
• litellm_v1.75.5-dev_memory_fix

Here's a snippet of my config.yaml:

  - model_name: gpt-5-nano
    litellm_params:
      model: azure/gpt-5-nano
      api_base: os.environ/AZURE_API_BASE # runs os.getenv("AZURE_API_BASE")
      api_key: os.environ/AZURE_API_KEY # runs os.getenv("AZURE_API_KEY")
      stream_timeout: 60
      merge_reasoning_content_in_choices: true
    model_info:
      mode: chat

Relevant log output

2025-08-13T19:03:39.2243278Z {"message": "litellm.router.Router::deployment_callback_on_success(): Exception occured - 'NoneType' object has no attribute 'startswith'", "level": "ERROR", "timestamp": "2025-08-13T19:03:39.218193", "stacktrace": "Traceback (most recent call last):\n  File \"/usr/lib/python3.13/site-packages/litellm/router.py\", line 4338, in deployment_callback_on_success\n    deployment_model_info = self.get_router_model_info(\n        deployment=deployment_info.model_dump(),\n        received_model_name=model_group,\n    )\n  File \"/usr/lib/python3.13/site-packages/litellm/router.py\", line 5477, in get_router_model_info\n    if not model.startswith(\"{}/\".format(custom_llm_provider)):\n           ^^^^^^^^^^^^^^^^\nAttributeError: 'NoneType' object has no attribute 'startswith'"}
2025-08-13T19:04:09.1131065Z {"message": "Could not identify azure model. Set azure 'base_model' for accurate max tokens, cost tracking, etc.- https://docs.litellm.ai/docs/proxy/cost_tracking#spend-tracking-for-azure-openai-models", "level": "ERROR", "timestamp": "2025-08-13T19:04:09.112900"}

Are you a ML Ops Team?

No

What LiteLLM version are you on ?

v1.75.5.dev3

Twitter / LinkedIn details

No response

[jugaldb]

Hi, please do not use dev releases, they are meant for checking certain things and might have older versions of the code, please try on the latest stable - https://github.com/BerriAI/litellm/releases/tag/v1.75.5.rc.1

do let us know if this continues to happen on the latest stable

[decent-engineer-decent-datascientist]

Apologies. I was intentional using dev to test the new ability to control reasoning effort. RC has many issues with Azure’s GPT 5 offerings. That being said, figured I’d raise this issue as I hadn’t seen it being discussed on the repo.

…

On Wed, Aug 13, 2025 at 4:03 PM Jugal D. Bhatt ***@***.***> wrote: *jugaldb* left a comment (BerriAI/litellm#13593) <#13593 (comment)> Hi, please do not use dev releases, they are meant for checking certain things and might have older versions of the code, please try on the latest stable - https://github.com/BerriAI/litellm/releases/tag/v1.75.5.rc.1 do let us know if this continues to happen on the latest stable — Reply to this email directly, view it on GitHub <#13593 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASRTZN6QVC3ILTNNLKN5VVD3NOR3PAVCNFSM6AAAAACD2R4UP2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCOBVHAYTCNZSGQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jayson0913]

10:23:36 - LiteLLM Proxy:ERROR: http_parsing_utils.py:87 - Unexpected error reading request body - Traceback (most recent call last): File "/usr/lib/python3.13/site-packages/litellm/proxy/common_utils/http_parsing_utils.py", line 40, in _read_request_body body = await request.body() ^^^^^^^^^^^^^^^^^^^^ File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 241, in body async for chunk in self.stream(): chunks.append(chunk) File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 235, in stream raise ClientDisconnect() starlette.requests.ClientDisconnect 10:23:36 - LiteLLM Proxy:ERROR: common_request_processing.py:642 - litellm.proxy.proxy_server._handle_llm_api_exception(): Exception occured - 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call /v1/modelsto view available models for your key.'} Traceback (most recent call last): File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 4055, in chat_completion result = await base_llm_response_processor.base_process_llm_request( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<16 lines>... ) ^ File "/usr/lib/python3.13/site-packages/litellm/proxy/common_request_processing.py", line 423, in base_process_llm_request llm_call = await route_request( ^^^^^^^^^^^^^^^^^^^^ ...<4 lines>... ) ^ File "/usr/lib/python3.13/site-packages/litellm/proxy/route_llm_request.py", line 170, in route_request raise ProxyModelNotFoundError( ...<2 lines>... ) litellm.proxy.route_llm_request.ProxyModelNotFoundError: 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call/v1/modelsto view available models for your key.'}
It seems that this issue still exists in the latest version, my version is v1.75.8-stable.
I started Litellm using Docker
The startup command is:
docker run -v ./config.yaml:/app/config.yaml -p 4000:4000 dockerhub.tokenfreeai.com/ghcrio/berriai/litellm:v1.75.8-stable --config /app/config.yaml

[buildgreatthings]

I have the same problem as well. It's not consistent, and somewhat randomly happens.

The LiteLLM logs are empty when the request happens, but on our client side we get a 502 from LiteLLM.

[krrishdholakia]

Does anyone have a consistent repro for this? Want to fix this.

[buildgreatthings]

Unfortunately not a consistent way to repo, but we found that it's likely due to the database refresh if the API key is still active. Ishaan suggested we bump our TTL key setting to 24 hours.

client log error:

"type": "InternalServerError",
"message": "502 status code (no body)",
"stack": "Error: 502 status code (no body) at APIError.generate (/app/node_modules/.pnpm/openai@6.3.0_ws@8.18.3_bufferutil@4.0.9__zod@3.25.76/node_modules/openai/src/core/error.ts:100:14) at OpenAI.makeStatusError (/app/node_modules/.pnpm/openai@6.3.0_ws@8.18.3_bufferutil@4.0.9__zod@3.25.76/node_modules/openai/src/client.ts:478:28) at OpenAI.makeRequest (/app/node_modules/.pnpm/openai@6.3.0_ws@8.18.3_bufferutil@4.0.9__zod@3.25.76/node_modules/openai/src/client.ts:728:24) at process.processTicksAndRejections (node:internal/process/task_queues:105:5) at async aiRunner (/app/src/server/utils/models/llm.ts:179:20) at async RetryOperation._fn (/app/src/server/utils/models/llm.ts:210:19)",
"status": 502, // 

error log on litellm

  "status": "failure",
  "batch_models": null,
  "usage_object": null,
  "user_api_key": "1cf6fbf92da97895f3a44c5d7f582b626dec0246da3924986df806692bd56a8a",
  "error_information": {
    "traceback": "  File \"/usr/lib/python3.13/site-packages/starlette/_exception_handler.py\", line 42, in wrapped_app\n    await app(scope, receive, sender)\n  File \"/usr/lib/python3.13/site-packages/starlette/routing.py\", line 73, in app\n    response = await f(request)\n               ^^^^^^^^^^^^^^^^\n  File \"/usr/lib/python3.13/site-packages/fastapi/routing.py\", line 291, in app\n    solved_result = await solve_dependencies(\n                    ^^^^^^^^^^^^^^^^^^^^^^^^^\n    ...<6 lines>...\n    )\n    ^\n  File \"/usr/lib/python3.13/site-packages/fastapi/dependencies/utils.py\", line 638, in solve_dependencies\n    solved = await call(**solved_result.values)\n             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/usr/lib/python3.13/site-packages/litellm/proxy/auth/user_api_key_auth.py\", line 1143, in user_api_key_auth\n    user_api_key_auth_obj = await _user_api_key_auth_builder(\n                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n    ...<8 lines>...\n    )\n    ^\n  File \"/usr/lib/python3.13/site-packages/litellm/proxy/auth/user_api_key_auth.py\", line 1110, in _user_api_key_auth_builder\n    return await UserAPIKeyAuthExceptionHandler._handle_authentication_error(\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n    ...<6 lines>...\n    )\n    ^\n  File \"/usr/lib/python3.13/site-packages/litellm/proxy/auth/auth_exception_handler.py\", line 118, in _handle_authentication_error\n    raise e\n  File \"/usr/lib/python3.13/site-packages/litellm/proxy/auth/user_api_key_auth.py\", line 964, in _user_api_key_auth_builder\n    raise ProxyException(\n    ...<4 lines>...\n    )\n",
    "error_code": "",
    "error_class": "ProxyException",
    "llm_provider": "",
    "error_message": ""
  },
  "applied_guardrails": null,
  "user_api_key_alias": null,
  "user_api_key_org_id": null,
  "user_api_key_team_id": null,
  "user_api_key_user_id": null,
  "guardrail_information": null,
  "model_map_information": null,
  "mcp_tool_call_metadata": null,
  "additional_usage_values": {},
  "user_api_key_team_alias": null,
  "vector_store_request_metadata": null

[krrishdholakia]

can you check your container logs for this? It might be an issue connecting to the db

[buildgreatthings]

Nealry identical errors with LiteLLM Logs:

21:30:30 - LiteLLM Proxy:ERROR: utils.py:882 - [Non-Blocking] Error in post_call_failure_hook: MyCustomHandler.async_post_call_failure_hook() got an unexpected keyword argument 'traceback_str'
    ...<4 lines>...
Traceback (most recent call last):
    )
  File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 3767, in chat_completion
    ^
    return await base_llm_response_processor.base_process_llm_request(
  File "/usr/lib/python3.13/site-packages/litellm/proxy/route_llm_request.py", line 166, in route_request
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    raise ProxyModelNotFoundError(
    ...<16 lines>...
    ...<2 lines>...
    )
    )
litellm.proxy.route_llm_request.ProxyModelNotFoundError: 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_request_processing.py", line 413, in base_process_llm_request
    llm_call = await route_request(
During handling of the above exception, another exception occurred:
           ^^^^^^^^^^^^^^^^^^^^
Traceback (most recent call last):
        ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 241, in body
  File "/usr/lib/python3.13/site-packages/litellm/proxy/utils.py", line 874, in post_call_failure_hook
    async for chunk in self.stream():
    _callback.async_post_call_failure_hook(
        chunks.append(chunk)
    ...<2 lines>...
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
        traceback_str=traceback_str,
        request_data=request_data,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
TypeError: MyCustomHandler.async_post_call_failure_hook() got an unexpected keyword argument 'traceback_str'
21:30:30 - LiteLLM Proxy:ERROR: http_parsing_utils.py:87 - Unexpected error reading request body - 
Traceback (most recent call last):
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_utils/http_parsing_utils.py", line 40, in _read_request_body
    body = await request.body()
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 235, in stream
    )
    raise ClientDisconnect()
starlette.requests.ClientDisconnect
21:30:30 - LiteLLM Proxy:ERROR: common_request_processing.py:630 - litellm.proxy.proxy_server._handle_llm_api_exception(): Exception occured - 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
Traceback (most recent call last):
  File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 3767, in chat_completion
    return await base_llm_response_processor.base_process_llm_request(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<16 lines>...
    )
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_request_processing.py", line 413, in base_process_llm_request
    llm_call = await route_request(
               ^^^^^^^^^^^^^^^^^^^^
    ...<4 lines>...
    )
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/route_llm_request.py", line 166, in route_request
    raise ProxyModelNotFoundError(
    ...<2 lines>...
litellm.proxy.route_llm_request.ProxyModelNotFoundError: 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_utils/http_parsing_utils.py", line 40, in _read_request_body
21:30:30 - LiteLLM Proxy:ERROR: http_parsing_utils.py:87 - Unexpected error reading request body - 
Traceback (most recent call last):
    body = await request.body()
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_utils/http_parsing_utils.py", line 40, in _read_request_body
           ^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 241, in body
    body = await request.body()
    async for chunk in self.stream():
           ^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 241, in body
    async for chunk in self.stream():
        chunks.append(chunk)
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 235, in stream
    raise ClientDisconnect()
starlette.requests.ClientDisconnect
21:30:30 - LiteLLM Proxy:ERROR: http_parsing_utils.py:87 - Unexpected error reading request body - 
Traceback (most recent call last):
        chunks.append(chunk)
  File "/usr/lib/python3.13/site-packages/starlette/requests.py", line 235, in stream
    raise ClientDisconnect()
starlette.requests.ClientDisconnect
21:30:30 - LiteLLM Proxy:ERROR: common_request_processing.py:630 - litellm.proxy.proxy_server._handle_llm_api_exception(): Exception occured - 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
Traceback (most recent call last):
  File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 3767, in chat_completion
    return await base_llm_response_processor.base_process_llm_request(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<16 lines>...
    )
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_request_processing.py", line 413, in base_process_llm_request
    llm_call = await route_request(
               ^^^^^^^^^^^^^^^^^^^^
    ...<4 lines>...
    )
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/route_llm_request.py", line 166, in route_request
    raise ProxyModelNotFoundError(
    ...<2 lines>...
litellm.proxy.route_llm_request.ProxyModelNotFoundError: 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
    )
litellm.proxy.route_llm_request.ProxyModelNotFoundError: 400: {'error': '/chat/completions: Invalid model name passed in model=None. Call `/v1/models` to view available models for your key.'}
21:30:30 - LiteLLM Proxy:ERROR: utils.py:882 - [Non-Blocking] Error in post_call_failure_hook: MyCustomHandler.async_post_call_failure_hook() got an unexpected keyword argument 'traceback_str'
Traceback (most recent call last):
  File "/usr/lib/python3.13/site-packages/litellm/proxy/proxy_server.py", line 3767, in chat_completion
    return await base_llm_response_processor.base_process_llm_request(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<16 lines>...
    )
    ^
  File "/usr/lib/python3.13/site-packages/litellm/proxy/common_request_processing.py", line 413, in base_process_llm_request
    llm_call = await route_request(
               ^^^^^^^^^^^^^^^^^^^^

[LoadingZhang]

met same problem, v1.80.0-stable