Deepseek API Usage with Langchain

In this article we’ll see how we can use the deepseek api from Lanchain framework as per documentation provided by Deepseek.

1. Install following library. Remove single quotes in command prompt.

        '!pip3 install langchain_openai'
   

2. Import BaseChatOpenAI class from the Langchain chat models

        from langchain_openai.chat_models.base import BaseChatOpenAI
   

3. Now create an BaseChatOpenAI object which will be used across our page.

        llm = BaseChatOpenAI(
            model='deepseek-chat', 
            openai_api_key='<Your API Key>'
            openai_api_base='https://api.deepseek.com',
            max_tokens=1024
        )
   

4. lets give an sample prompt message to check whether the api is working fine.

        response = llm.invoke("Hi!")
        print(response.content)
   

        Hello! How can I assist you today? 
   

5. lets givea proper message and see the output to ensure that API is responding as expected

        messages = [
            (
                "system",
                "You are a helpful assistant that translates English to Persian. Translate the user sentence.",
            ),
            ("human", "I love programming."),
        ]
        ai_msg = llm.invoke(messages)
        ai_msg.content
   

6. Check with complete message given to system

       
        from langchain_core.prompts import ChatPromptTemplate
           
        prompt = ChatPromptTemplate(
            [
                (
                    "system",
                    "You are a helpful assistant that translates {input_language} to {output_language}.",
                ),
                ("human", "{input}"),
            ]
        )
           
        chain = prompt | llm
        chain.invoke(
            {
                "input_language": "English",
                "output_language": "German",
                "input": "I love agentic AI.",
            }
        )
   

7. Observe the output, esepcailly the KV Cache or content caching portion.

        AIMessage(content='Ich liebe agentische KI.', additional_kwargs={'refusal': None}, response_metadata={'token_usage': {'completion_tokens': 7, 'prompt_tokens': 20, 'total_tokens': 27, 'completion_tokens_details': None, 'prompt_tokens_details': {'audio_tokens': None, 'cached_tokens': 0}, 'prompt_cache_hit_tokens': 0, 'prompt_cache_miss_tokens': 20}, 'model_name': 'deepseek-chat', 'system_fingerprint': 'fp_3a5770e1b4', 'finish_reason': 'stop', 'logprobs': None}, id='run-46ca6f2a-21ef-45f4-83f6-814c07fab391-0', usage_metadata={'input_tokens': 20, 'output_tokens': 7, 'total_tokens': 27, 'input_token_details': {'cache_read': 0}, 'output_token_details': {}})
   

8. You can find working google colab file in my github repo.