Educhain is a powerful Python package that leverages Generative AI to create engaging and personalized educational content. From generating multiple-choice questions to crafting comprehensive lesson plans, Educhain makes it easy to apply AI in various educational scenarios.
📝 Generate Multiple Choice Questions (MCQs)
from educhain import Educhain
client = Educhain()
# Basic MCQ generation
mcq = client.qna_engine.generate_questions(
topic="Solar System",
num=3,
question_type="Multiple Choice"
)
# Advanced MCQ with custom parameters
advanced_mcq = client.qna_engine.generate_questions(
topic="Solar System",
num=3,
question_type="Multiple Choice",
difficulty_level="Hard",
custom_instructions="Include recent discoveries"
)
print(mcq.model_dump_json()) # View in JSON format , For Dictionary format use mcq.model_dump()📊 Create Lesson Plans
from educhain import Educhain
client = Educhain()
# Basic lesson plan
lesson = client.content_engine.generate_lesson_plan(
topic="Photosynthesis"
)
# Advanced lesson plan with specific parameters
detailed_lesson = client.content_engine.generate_lesson_plan(
topic="Photosynthesis",
duration="60 minutes",
grade_level="High School",
learning_objectives=["Understanding the process", "Identifying key components"]
)
print(lesson.model_dump_json()) # View in JSON format , For Dictionary format use lesson.model_dump()🔄 Support for Various LLM Models
from educhain import Educhain, LLMConfig
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_openai import ChatOpenAI
# Using Gemini
gemini_model = ChatGoogleGenerativeAI(
model="gemini-2.0-flash",
google_api_key="YOUR_GOOGLE_API_KEY"
)
gemini_config = LLMConfig(custom_model=gemini_model)
gemini_client = Educhain(gemini_config)
# Using GPT-4
gpt4_model = ChatOpenAI(
model_name="gpt-4.1",
openai_api_key="YOUR_OPENAI_API_KEY"
)
gpt4_config = LLMConfig(custom_model=gpt4_model)
gpt4_client = Educhain(gpt4_config)📁 Export Questions to Different Formats
from educhain import Educhain
client = Educhain()
questions = client.qna_engine.generate_questions(topic="Climate Change", num=5)
# Export to JSON
questions.json("climate_questions.json")
# Export to PDF
questions.to_pdf("climate_questions.pdf")
# Export to CSV
questions.to_csv("climate_questions.csv")🎨 Customizable Prompt Templates
from educhain import Educhain
client = Educhain()
# Custom template for questions
custom_template = """
Generate {num} {question_type} questions about {topic}.
Ensure the questions are:
- At {difficulty_level} level
- Focus on {learning_objective}
- Include practical examples
- {custom_instructions}
"""
questions = client.qna_engine.generate_questions(
topic="Machine Learning",
num=3,
question_type="Multiple Choice",
difficulty_level="Intermediate",
learning_objective="Understanding Neural Networks",
custom_instructions="Include recent developments",
prompt_template=custom_template
)📚 Generate Questions from Files
from educhain import Educhain
client = Educhain()
# From URL
url_questions = client.qna_engine.generate_questions_from_data(
source="https://example.com/article",
source_type="url",
num=3
)
# From PDF
pdf_questions = client.qna_engine.generate_questions_from_data(
source="path/to/document.pdf",
source_type="pdf",
num=3
)
# From Text File
text_questions = client.qna_engine.generate_questions_from_data(
source="path/to/content.txt",
source_type="text",
num=3
)📹 Generate Questions from YouTube Videos 
from educhain import Educhain
client = Educhain()
# Basic usage - Generate 3 MCQs from a YouTube video
questions = client.qna_engine.generate_questions_from_youtube(
url="https://www.youtube.com/watch?v=dQw4w9WgXcQ",
num=3
)
print(questions.model_dump_json())
# Generate questions preserving original language
preserved_questions = client.qna_engine.generate_questions_from_youtube(
url="https://www.youtube.com/watch?v=dQw4w9WgXcQ",
num=2,
target_language='hi',
preserve_original_language=True # Keeps original language
)🥽 Generate Questions from Images
from educhain import Educhain
client = Educhain() #Default is 4o-mini (make sure to use a multimodal LLM!)
question = client.qna_engine.solve_doubt(
image_source="path-to-your-image",
prompt="Explain the diagram in detail",
detail_level = "High"
)
print(question)🥽 Generate Visual Questions
from langchain_google_genai import ChatGoogleGenerativeAI
from educhain import Educhain, LLMConfig
gemini_flash = ChatGoogleGenerativeAI(model="gemini-2.0-flash", google_api_key=GOOGLE_API_KEY)
flash_config = LLMConfig(custom_model=gemini_flash)
client = Educhain(flash_config)
ques = client.qna_engine.generate_visual_questions(
topic="GMAT Statistics", num=10 )
print(ques.model_dump_json())Reimagining Education with AI 🤖
- 📜 QnA Engine: Generates an infinte variety of Questions
- 📰 Content Engine: One-stop content generation - lesson plans, flashcards, notes etc
- 📌 Personalization Engine: Adapts to your individual level of understanding for a tailored experience.
pip install educhain
Get started with content generation in < 3 lines!
from educhain import Educhain
client = Educhain()
ques = client.qna_engine.generate_questions(topic="Newton's Law of Motion",
num=5)
print(ques.model_dump_json())
ques.model_dump_json() # ques.model_dump()Generates different types of questions. See the advanced guide to create a custom question type.
# Supports "Multiple Choice" (default); "True/False"; "Fill in the Blank"; "Short Answer"
from educhain import Educhain
client = Educhain()
ques = client.qna_engine.generate_questions(topic = "Psychology",
num = 10,
question_type="Fill in the Blank"
custom_instructions = "Only basic questions")
print(ques.model_dump_json())
ques.model_dump_json() #ques.model_dump()To use a custom model, you can pass a model configuration through the LLMConfig class
Here's an example using the Gemini Model
from langchain_google_genai import ChatGoogleGenerativeAI
from educhain import Educhain, LLMConfig
gemini_flash = ChatGoogleGenerativeAI(
model="gemini-2.0-flash",
google_api_key="GOOGLE_API_KEY")
flash_config = LLMConfig(custom_model=gemini_flash)
client = Educhain(flash_config) #using gemini model with educhain
ques = client.qna_engine.generate_questions(topic="Psychology",
num=10)
print(ques.model_dump_json())
ques.model_dump_json() #ques.model_dump()Configure your prompt templates for more control over input parameters and output quality.
from educhain import Educhain
client = Educhain()
custom_template = """
Generate {num} multiple-choice question (MCQ) based on the given topic and level.
Provide the question, four answer options, and the correct answer.
Topic: {topic}
Learning Objective: {learning_objective}
Difficulty Level: {difficulty_level}
"""
ques = client.qna_engine.generate_questions(
topic="Python Programming",
num=2,
learning_objective="Usage of Python classes",
difficulty_level="Hard",
prompt_template=custom_template,
)
print(ques.model_dump_json())Ingest your own data to create content. Currently supports URL/PDF/TXT.
from educhain import Educhain
client = Educhain()
ques = client.qna_engine.generate_questions_from_data(
source="https://en.wikipedia.org/wiki/Big_Mac_Index",
source_type="url",
num=5)
print(ques.model_dump_json())
ques.model_dump_json() # ques.model_dump()Create interactive and detailed lesson plans.
from educhain import Educhain
client = Educhain()
plan = client.content_engine.generate_lesson_plan(
topic = "Newton's Law of Motion")
print(plan.model_dump_json())
plan.model_dump_json() # plan.model_dump()- Multiple Choice Questions (MCQ)
- Short Answer Questions
- True/False Questions
- Fill in the Blank Questions
API Key Authentication Errors
Error: Authentication failed. Please check your API key.
Solution: Verify that your API key is correct and properly set. For OpenAI or Google API keys, ensure they are active and have sufficient quota remaining.
# Correct way to set API keys
import os
os.environ["OPENAI_API_KEY"] = "your-api-key-here"
# or
os.environ["GOOGLE_API_KEY"] = "your-api-key-here"Model Not Generating Expected Output
Issue: The model generates content that doesn't match your expectations or requirements.
Solution: Try adjusting the parameters or providing more specific instructions:
# Be more specific with your requirements
questions = client.qna_engine.generate_questions(
topic="Python Programming",
num=3,
difficulty_level="Intermediate",
custom_instructions="Focus on object-oriented programming concepts. Include code examples in each question."
)Package Import Errors
ModuleNotFoundError: No module named 'educhain'
Solution: Ensure you've installed the package correctly:
pip install educhain --upgradeIf you're using a virtual environment, make sure it's activated before installing.
Memory Issues with Large Outputs
Issue: Generating a large number of questions causes memory errors.
Solution: Generate questions in smaller batches:
# Instead of generating 50 questions at once
all_questions = []
for i in range(5):
batch = client.qna_engine.generate_questions(
topic="History",
num=10
)
all_questions.extend(batch.questions)- Bulk Generation
- Outputs in JSON format
- Custom Prompt Templates
- Custom Response Models using Pydantic
- Exports questions to JSON/PDF/CSV
- Support for other LLM models
- Generate questions from text/PDF file
- Integration with popular Learning Management Systems
- Mobile app for on-the-go content generation
We invite you to help enhance our library. If you have any ideas, improvements, or suggestions for enhancements to contribute, please open a GitHub issue or submit a pull request. Be sure to adhere to our existing project structure and include a detailed README.md for any new Contribution.
Thank you for your continued support, community!
- ✨ Added support for generating visual questions with multimodal LLMs
- ✨ Added support for generating questions from YouTube videos
- ✨ Added support for generating questions from images
- 🐛 Fixed issue with PDF parsing for certain file formats
- ⚡️ Improved performance for large document processing
- ✨ Added support for custom prompt templates
- ✨ Added export functionality to PDF, CSV, and JSON
- 🔄 Enhanced compatibility with Gemini models
- 📚 Expanded documentation with more examples
- 🚀 Initial release
- ✅ Core question generation functionality
- ✅ Support for multiple question types
- ✅ Basic lesson plan generation
- ✅ Integration with OpenAI models
This project is licensed under the MIT License - see the LICENSE file for details.
Made with ❤️ by Buildfastwithai
