OpenAIEmbeddings unable to embed_query or embed_documents to locally hosted embedding but works directly.

[kevin-liangit]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain documentation with the integrated search.
• 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).

Example Code

from langchain_openai import ChatOpenAI,OpenAIEmbeddings

embeddings_model = OpenAIEmbeddings(model="bge-embed-text-cpu",base_url="http://localhost:8000/openai/v1",api_key="thisIsIgnored",deployment="bge-embed-text-cpu")

Error Message and Stack Trace (if applicable)

Errors when using embeddings_model.embed_query

embeddings_model.embed_query("test")
Traceback (most recent call last):
File "", line 1, in
File "/lib/python3.11/site-packages/langchain_openai/embeddings/base.py", line 629, in embed_query
return self.embed_documents([text])[0]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lib/python3.11/site-packages/langchain_openai/embeddings/base.py", line 588, in embed_documents
return self._get_len_safe_embeddings(texts, engine=engine)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lib/python3.11/site-packages/langchain_openai/embeddings/base.py", line 483, in _get_len_safe_embeddings
response = self.client.create(
^^^^^^^^^^^^^^^^^^^
File "/lib/python3.11/site-packages/openai/resources/embeddings.py", line 124, in create
return self._post(
^^^^^^^^^^^
File "/lib/python3.11/site-packages/openai/_base_client.py", line 1278, in post
return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lib/python3.11/site-packages/openai/_base_client.py", line 955, in request
return self._request(
^^^^^^^^^^^^^^
File "/lib/python3.11/site-packages/openai/_base_client.py", line 1059, in _request
raise self._make_status_error_from_response(err.response) from None
openai.UnprocessableEntityError: Error code: 422 - {'detail': [{'type': 'string_type', 'loc': ['body', 'text', 'input', 'list[constrained-str]', 0], 'msg': 'Input should be a valid string', 'input': [1985]}, {'type': 'string_type', 'loc': ['body', 'text', 'input', 'constrained-str'], 'msg': 'Input should be a valid string', 'input': [[1985]]}]}

Description

Unable to use OpenAIEmbeddings with a locally hosted embedding.

The following does work when using the client directly

embeddings_model.client.create(input="Your text goes here.",model="bge-embed-text-cpu")
CreateEmbeddingResponse(data=[Embedding(embedding=[-0.04428496211767197, -0.029685525223612785, 0.028833890333771706, -0.03357870876789093, -0.024697383865714073, -0.05158468335866928, 0.07737702131271362, 0.06959065794944763, 0.03966180980205536, -0.04136507585644722, -0.00024047252372838557, -0.028590567409992218, 0.016181044280529022, -0.011922874487936497, 0.02092586271464824.....

Also can confirm through openai directly.

from openai import OpenAI
Assumes port-forward of kubeai service to localhost:8000.
client = OpenAI(api_key="ignored", base_url="http://localhost:8000/openai/v1")
response = client.embeddings.create(
    input="Your text goes here.",
    model="bge-embed-text-cpu"
)

Any ideas?

System Info

langchain==0.3.11
langchain-community==0.3.11
langchain-core==0.3.24
langchain-google-community==2.0.2
langchain-google-vertexai==2.0.3
langchain-huggingface==0.1.2
langchain-openai==0.2.10
langchain-text-splitters==0.3.2
langgraph==0.2.53
langgraph-checkpoint==2.0.6
langgraph-sdk==0.1.36
langsmith==0.1.146

[kevin-liangit]

seems to be related when using engine: infinity

michaelfeil/infinity#36

using IninfityEmbeddings works, as outlined in michaelfeil/infinity#329

embeddings_model = InfinityEmbeddings(model="bge-embed-text-cpu",infinity_api_url="http://localhost:8000/openai/v1")

[dosubot]

Hi, @kevin-liangit. I'm Dosu, and I'm helping the LangChain team manage their backlog. I'm marking this issue as stale.

Issue Summary:

• You reported a bug with the OpenAIEmbeddings class failing to embed queries/documents using a locally hosted model.
• The issue is linked to using the engine "infinity," as noted in a related issue on the Infinity repository.
• You found a workaround by using InfinityEmbeddings with a specific configuration, which resolves the problem.

Next Steps:

• Could you confirm if this issue is still relevant with the latest version of LangChain? If so, please comment to keep the discussion open.
• If there are no further updates, this issue will be automatically closed in 7 days.

Thank you for your understanding and contribution!