When passing api_key and not ad_token, condition takes the empty string of the ad token instead of apo_key

[erne2086]

Checked other resources

• I added a very descriptive title to this issue.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).
• I posted a self-contained, minimal, reproducible example. A maintainer can copy it and run it AS IS.

Example Code

The following code reproduces the error:

import os
#change for your details
os.environ["AZURE_AI_SEARCH_SERVICE_NAME"] = "service_name"
os.environ["AZURE_AI_SEARCH_INDEX_NAME"] = "index_name"
os.environ["AZURE_AI_SEARCH_API_KEY"] = "your-api-key"
os.environ["AZURE_AI_SEARCH_ENDPOINT"] = "your-endpoint"

from langchain_community.retrievers import AzureAISearchRetriever

retriever = AzureAISearchRetriever(
    content_key="content", 
    top_k=10, 
    index_name="your-index", 
    service_name="your-service",
)

retriever.invoke("What is the capital of France?")

Error Message and Stack Trace (if applicable)

---

Exception Traceback (most recent call last)
Cell In[40], line 1
----> 1 retriever.invoke("What is the capital of France?")

File ~/Documents/coding/agentic-frameworks/langchain-demo/.venv/lib/python3.13/site-packages/langchain_core/retrievers.py:258, in BaseRetriever.invoke(self, input, config, **kwargs)
256 _kwargs = kwargs if self._expects_other_args else {}
257 if self._new_arg_supported:
--> 258 result = self._get_relevant_documents(
259 input, run_manager=run_manager, **_kwargs
260 )
261 else:
262 result = self._get_relevant_documents(input, **_kwargs)

File ~/Documents/coding/agentic-frameworks/langchain-demo/.venv/lib/python3.13/site-packages/langchain_community/retrievers/azure_ai_search.py:206, in AzureAISearchRetriever._get_relevant_documents(self, query, run_manager)
203 def _get_relevant_documents(
204 self, query: str, *, run_manager: CallbackManagerForRetrieverRun
205 ) -> List[Document]:
--> 206 search_results = self._search(query)
208 return [
209 Document(page_content=result.pop(self.content_key), metadata=result)
210 for result in search_results
211 ]

File :6, in _search(self, query)

Exception: Error in search request: <Response [403]>

Description

I am trying to use the AzureAISearchRetriever from langchain_community.retrievers on langchain-community

When I want to authenticate with api-key, I get a 403 error, I traced back the error to this line:

I modified the code as mentioned in this issue:

langchain-ai/langchain@68d16d8#r155860761

And I was able to authenticate with api_key.

System Info

uv pip show langchain-community
Name: langchain-community
Version: 0.3.22
Location: /Users/ejcv/Documents/coding/agentic-frameworks/langchain-demo/.venv/lib/python3.13/site-packages
Requires: aiohttp, dataclasses-json, httpx-sse, langchain, langchain-core, langsmith, numpy, pydantic-settings, pyyaml, requests, sqlalchemy, tenacity
Required-by:

uv pip show langchain          
Name: langchain
Version: 0.3.24
Location: /Users/ejcv/Documents/coding/agentic-frameworks/langchain-demo/.venv/lib/python3.13/site-packages
Requires: langchain-core, langchain-text-splitters, langsmith, pydantic, pyyaml, requests, sqlalchemy
Required-by: langchain-community

[kavinkumar807]

@erne2086 will check with this

[tol-dawn]

I experienced this issue as well, same versions as described above. If you don't want to modify the code in langchain, a quick workaround can be to set some abitrary value for the "azure_ad_token" argument. Like this:

retriever = AzureAISearchRetriever(
                content_key="content",
                top_k=3,
                index_name=search_index_name,
                service_name=ai_search_endpoint,
                azure_ad_token="True")

[Satya-Indukuri]

Hi! I'm working on this issue as part of my university open-source contribution project. I plan to submit a PR shortly. Please let me know if there's anything I should consider before proceeding. Thanks!