Chat with PDF is a free, interactive chatbot that allows you to upload PDF documents and ask questions directly about their content. The chatbot uses state-of-the-art open-source language models to extract, embed, and retrieve relevant information, making your PDFs searchable and conversational.
Upload multiple PDF documents at once.
Extracts text from PDFs and splits it into meaningful chunks.
Uses semantic embeddings for accurate context retrieval.
Conversational interface remembers chat history.
Provides precise answers based on PDF content only.
Fully free and open-source, no paid APIs required.
PDF Processing: Extracts text from uploaded PDFs and cleans it.
Text Chunking: Splits text into manageable chunks for better context handling.
Vector Embedding: Converts chunks into vector embeddings using open-source models.
Semantic Search: Retrieves relevant chunks based on user queries.
Conversational AI: Generates accurate responses using HuggingFace’s T5 models.
Memory: Maintains conversation history for a smooth chat experience.
Frontend: Streamlit
PDF Processing: PyPDF2
Text Embeddings & Retrieval: LangChain + HuggingFace Embeddings + FAISS
Language Model: HuggingFace google/flan-t5-base
Memory Management: LangChain ConversationBufferMemory
Python Libraries: Transformers, LangChain, Streamlit