Make the OpenAI LLM provider adapt to OpenAI-compatible servers and reasoning models. #10570

docs/en_US/ai_tools.rst

@@ -56,7 +56,8 @@ Select your preferred LLM provider from the dropdown:
 
 **OpenAI**
   Use GPT models from OpenAI, or any OpenAI-compatible API provider (e.g.,
-  LiteLLM, LM Studio, EXO, or other local inference servers).
+  LiteLLM, LM Studio, EXO, hosted services such as GLM (Z.ai), or local
+  inference servers).
 
   * **API URL**: Custom API endpoint URL (leave empty for default: https://api.openai.com/v1). Include the ``/v1`` path prefix if required by your provider.
   * **API Key File**: Path to a file containing your OpenAI API key (obtain from https://platform.openai.com/). This path refers to the filesystem where the pgAdmin server is running (e.g., inside the container if using Docker). The ``~`` prefix is expanded to the home directory of the user running the pgAdmin server process. Optional when using a custom URL with a provider that does not require authentication.
@@ -77,7 +78,10 @@ Select your preferred LLM provider from the dropdown:
 
 .. note:: You can also use the *OpenAI* provider with a custom API URL for any
    OpenAI-compatible endpoint, including Docker Model Runner and other local
-   inference servers.
+   inference servers. pgAdmin sends ``max_tokens`` by default and learns to use
+   ``max_completion_tokens`` when a server explicitly rejects it. When the
+   ``ALLOWED_LLM_API_URLS`` configuration setting is non-empty, custom API URLs must match one of its entries; an
+   empty list permits any URL.
 
 After configuring your provider, click *Save* to apply the changes.
 

docs/en_US/preferences.rst

@@ -66,7 +66,8 @@ Use the fields on the *AI* panel to configure your LLM provider:
 
 * Use the *API URL* field to set a custom API endpoint URL. Leave empty to use
   the default OpenAI API (``https://api.openai.com/v1``). Set a custom URL to
-  use any OpenAI-compatible API provider (e.g., LiteLLM, LM Studio, EXO).
+  use any OpenAI-compatible API provider (e.g., LiteLLM, LM Studio, EXO,
+  GLM (Z.ai)).
   Include the ``/v1`` path prefix if required by your provider
   (e.g., ``http://localhost:1234/v1``).
 
@@ -101,8 +102,14 @@ Use the fields on the *AI* panel to configure your LLM provider:
 
 .. note:: You can also use the *OpenAI* provider with a custom API URL for any
    OpenAI-compatible endpoint, including Docker Model Runner, LM Studio, EXO,
-   and other local inference servers. This can be useful when you want to use
-   a provider that isn't explicitly listed but supports the OpenAI API format.
+   hosted services such as GLM (Z.ai), and other local inference servers. This
+   can be useful when you want to use a provider that isn't explicitly listed
+   but supports the OpenAI API format.
+
+.. note:: With the default allowlist enabled, an administrator must add any
+   API host not already in ``ALLOWED_LLM_API_URLS`` before pgAdmin can connect
+   to it. By default the list allows the OpenAI and Anthropic APIs and loopback
+   addresses on any port. An empty allowlist disables this restriction.
 
 The Browser Node
 ****************

web/pgadmin/llm/models.py

@@ -194,6 +194,8 @@ class LLMError:
     code: Optional[str] = None
     provider: Optional[str] = None
     retryable: bool = False
+    http_status: Optional[int] = None
+    param: Optional[str] = None
 
     def __str__(self) -> str:
         if self.code: