docs/integrations/chat/deepseek/

[giscus[bot]]

docs/integrations/chat/deepseek/

This will help you getting started with DeepSeek's hosted chat models. For detailed documentation of all ChatDeepSeek features and configurations head to the API reference.

https://python.langchain.com/docs/integrations/chat/deepseek/

[rahulbhoyar1995]

Whenever I am trying to install I am getting this error.

ERROR: Could not find a version that satisfies the requirement langchain-deepseek-official (from versions: none)
ERROR: No matching distribution found for langchain-deepseek-official

[ccurme]

Hi @rahulbhoyar1995, from your Github issue it sounds like you are running python 3.8. This package requires python 3.9+.

[LErdweg]

I am unable to add this package to my LangGraph app because it causes unsolvable package conflicts with pip dependencies. I've tried resolving the conflicts, but pip fails to find a compatible set of dependencies. Are there any workarounds or alternative approaches to make this package work within my LangGraph setup?

[ccurme]

What errors are you seeing?

[LErdweg]

We can install and compile all of our other packages successfully. However, once we add langchain-deepseek-official into requirements.in, pip-compile fails with unsolvable dependency constraints. Specifically:

langchain-deepseek-official pins versions of langchain-core and langchain-openai that conflict with other LangChain libraries (e.g., langchain-postgres, langchain, langchain-community, etc.) we need.
We get a ResolutionImpossible error in pip because no version combination of langchain-openai and langchain-core can satisfy all pinned constraints from langchain-deepseek-official alongside our other packages.
Once we remove langchain-deepseek-official, the pip-compile process works without errors.

[cugbomah]

Any timelines on support for running this model locally?

[cugbomah]

To clarify a point, when ran locally using Ollama, I get an error to the effect that the model does not support tools.

[feelfree0928]

Hi , I am developing a chatbot using Deepseek, so I used langchain_deepseek package.

../langchain_deepseek/chat_models.py (Importing openai package in deepseek package)
"""DeepSeek chat models."""
from typing import Dict, Optional, Type, Union
import openai
...
from langchain_openai.chat_models.base import BaseChatOpenAI

That's exactly why I guess the error below occurs.
...
File "/root/resume-ranking/analysis_service/venv/lib/python3.11/site-packages/langchain_openai/chat_models/base.py", line 792, in _generate
response = self.client.create(**payload)
openai.APIStatusError: Error code: 402 - {'error': {'message': 'Insufficient Balance', 'type': 'unknown_error', 'param': None, 'code': 'invalid_request_error'}}
...

how can I fix this?

[feelfree0928]

I found that Deepseek was much slower than I expected and was disappointed. Now, I don't need Deepseek!
Thanks!

[Nawinkumar97]

So deepSeek doesn't support embedding so how can I integrate Deepseek API for embeddings