🧠 Django + LangChain End-to-End Cheatsheet (AI-Ready Backend)

Use this cheat sheet to integrate a LangChain-powered AI agent with your Django app.

---

🔧 1. Install Dependencies

pip install django langchain openai python-dotenv

---

🧠 LangChain LLM Setup (OpenAI, Gemini, Claude, DeepSeek, Mistral)

Install the appropriate packages depending on the model provider you want to use:

Provider	Installation Command
OpenAI	pip install openai langchain
Gemini	pip install langchain-google-genai
Claude	pip install anthropic langchain
DeepSeek	pip install langchain openai (used via OpenAI-compatible APIs)
Mistral / Mixtral	pip install langchain huggingface_hub or use together API

---

🔌 LangChain Setup Per Provider

✅ OpenAI

from langchain.chat_models import ChatOpenAI
llm = ChatOpenAI(openai_api_key=os.getenv("OPENAI_API_KEY"))

✅ Gemini (Google)

from langchain_google_genai import ChatGoogleGenerativeAI
llm = ChatGoogleGenerativeAI(model="gemini-pro", google_api_key=os.getenv("GOOGLE_API_KEY"))

✅ Claude (Anthropic)

from langchain.chat_models import ChatAnthropic
llm = ChatAnthropic(anthropic_api_key=os.getenv("ANTHROPIC_API_KEY"), model_name="claude-3-sonnet-20240229")

✅ DeepSeek (via OpenAI-compatible endpoint)

from langchain.chat_models import ChatOpenAI
llm = ChatOpenAI(
    openai_api_key=os.getenv("DEEPSEEK_API_KEY"),
    base_url="https://api.deepseek.com/v1",
    model="deepseek-chat"
)

✅ Mistral/Mixtral (via Hugging Face or Together API)

from langchain.llms import HuggingFaceHub
llm = HuggingFaceHub(repo_id="mistralai/Mixtral-8x7B-Instruct-v0.1", huggingfacehub_api_token=os.getenv("HF_API_KEY"))

Or via Together:

from langchain.chat_models import ChatOpenAI
llm = ChatOpenAI(
    openai_api_key=os.getenv("TOGETHER_API_KEY"),
    base_url="https://api.together.xyz/v1",
    model="mistralai/Mixtral-8x7B-Instruct-v0.1"
)

---

📁 2. Project Structure

myproject/
├── myapp/
│   ├── views.py
│   ├── urls.py
│   └── ...
├── myproject/
│   └── settings.py
├── .env
└── manage.py

---

🔐 3. .env Configuration

OPENAI_API_KEY=your_openai_key
GOOGLE_API_KEY=your_gemini_key

In settings.py:

import os
from dotenv import load_dotenv
load_dotenv()

---

⚙️ 4. LangChain Setup (e.g., Chatbot)

# myapp/ai/agent.py
from langchain.chat_models import ChatOpenAI
from langchain.prompts import ChatPromptTemplate
from langchain.chains import LLMChain
import os

def build_chain():
    llm = ChatOpenAI(api_key=os.getenv("OPENAI_API_KEY"), temperature=0.7)
    prompt = ChatPromptTemplate.from_template("You are an Islamic insurance advisor. Q: {question} A:")
    chain = LLMChain(llm=llm, prompt=prompt)
    return chain

def get_ai_response(question):
    chain = build_chain()
    return chain.run({"question": question})

---

🌐 5. Django View for AI Endpoint

# myapp/views.py
from django.http import JsonResponse
from .ai.agent import get_ai_response

def ai_chat(request):
    question = request.GET.get("q", "")
    if not question:
        return JsonResponse({"error": "No question provided"}, status=400)
    
    response = get_ai_response(question)
    return JsonResponse({"response": response})

---

🔗 6. Add URL Route

# myapp/urls.py
from django.urls import path
from .views import ai_chat

urlpatterns = [
    path("chat/", ai_chat, name="ai_chat"),
]

And include in myproject/urls.py:

from django.urls import path, include

urlpatterns = [
    path("api/", include("myapp.urls")),
]

---

⚡ 7. Run and Query

python manage.py runserver

Query Example:

GET http://localhost:8000/api/chat/?q=What is ..

---

🧠 Optional: Use Gemini (Instead of OpenAI)

from langchain_google_genai import ChatGoogleGenerativeAI

llm = ChatGoogleGenerativeAI(model="gemini-pro", google_api_key=os.getenv("GOOGLE_API_KEY"))

---

✅ Tips

• Use LangChain Expression Language (LCEL) for composability.
• Wrap LangChain in async Django views for performance.
• Add semantic memory with FAISS, ChromaDB, or Weaviate.
• For production: add CORS, rate limiting, auth (e.g., JWT), and async views.