[Bug]: Cannot import from llama_index.llms.langchain import LangChainLLM

[rajeshkochi444]

Bug Description

cannot import from llama_index.llms.langchain import LangChainLLM and cannot use llm = LangChainLLM(llm=llm_lc)

---

ModuleNotFoundError Traceback (most recent call last)
Cell In[8], line 1
----> 1 from llama_index.llms.langchain import LangChainLLM
2 from llama_index.embeddings.langchain import LangchainEmbedding
4 llm = LangChainLLM(llm=llm_lc)

File ~/.venv/llm_rag/lib/python3.10/site-packages/llama_index/llms/langchain/init.py:1
----> 1 from llama_index.llms.langchain.base import LangChainLLM
3 all = ["LangChainLLM"]

File ~/.venv/llm_rag/lib/python3.10/site-packages/llama_index/llms/langchain/base.py:23
20 from llama_index.core.llms.llm import LLM
21 from llama_index.core.types import BaseOutputParser, PydanticProgramMode, Thread
---> 23 from langchain.base_language import BaseLanguageModel
24 from langchain.schema import AIMessage
27 class LangChainLLM(LLM):

ModuleNotFoundError: No module named 'langchain.base_language'

Version

0.14.9

Steps to Reproduce

from llama_index.llms.langchain import LangChainLLM

---

ModuleNotFoundError Traceback (most recent call last)
Cell In[8], line 1
----> 1 from llama_index.llms.langchain import LangChainLLM
2 from llama_index.embeddings.langchain import LangchainEmbedding
4 llm = LangChainLLM(llm=llm_lc)

File ~/.venv/llm_rag/lib/python3.10/site-packages/llama_index/llms/langchain/init.py:1
----> 1 from llama_index.llms.langchain.base import LangChainLLM
3 all = ["LangChainLLM"]

File ~/.venv/llm_rag/lib/python3.10/site-packages/llama_index/llms/langchain/base.py:23
20 from llama_index.core.llms.llm import LLM
21 from llama_index.core.types import BaseOutputParser, PydanticProgramMode, Thread
---> 23 from langchain.base_language import BaseLanguageModel
24 from langchain.schema import AIMessage
27 class LangChainLLM(LLM):

ModuleNotFoundError: No module named 'langchain.base_language'

Relevant Logs/Tracbacks

[dosubot]

Hi @rajeshkochi444! I'm Dosu and I’m helping the llama_index team.

The import error happens because LangChainLLM expects to import BaseLanguageModel from langchain.base_language, but this module was removed in LangChain 1.x and later. To fix this, check your installed LangChain version with pip show langchain or pip freeze. If you have LangChain 1.x or newer, downgrade to a compatible version like 0.3.x using:

pip install langchain==0.3.0

There’s currently no migration guide or patch for LangChain 1.x support in llama-index-llms-langchain. If you need LangChain 1.x features, consider using other LlamaIndex integrations (like llama-index-llms-ollama) or writing a custom adapter. For more details, see this issue thread and the implementation reference.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[logan-markewich]

ah yea, this isn't updated to support langchain 1.0 yet. Is there a specific reason you need a langchain LLM? Things generally work better when using the llama-index LLMs

[rajeshkochi444]

for some reason I cannot use

embed_model = AzureOpenAIEmbedding(
model=os.getenv("AZURE_EMBED_DEPLOYMENT_NAME"),
deployment_name=os.getenv("AZURE_EMBED_DEPLOYMENT_NAME"),
api_key=os.getenv("AZURE_EMBED_API_KEY"),
azure_endpoint=os.getenv("AZURE_EMBED_ENDPOINT"),
api_version=os.getenv("AZURE_OPENAI_API_VERSION"),
)

[logan-markewich]

@rajeshkochi444 model and deployment name are typically not the same thing. The model should be the actual embedding model name (text-embedding-ada-002 or text-embedding-3-small for example)

api_key = "<api-key>"
azure_endpoint = "https://<your-resource-name>.openai.azure.com/"
api_version = "2023-07-01-preview"

llm = AzureOpenAI(
    model="gpt-35-turbo-16k",
    deployment_name="my-custom-llm",
    api_key=api_key,
    azure_endpoint=azure_endpoint,
    api_version=api_version,
)

# You need to deploy your own embedding model as well as your own chat completion model
embed_model = AzureOpenAIEmbedding(
    model="text-embedding-ada-002",
    deployment_name="my-custom-embedding",
    api_key=api_key,
    azure_endpoint=azure_endpoint,
    api_version=api_version,
)