Enhances Tesseract OCR output using LLMs (local or API) for error correction, smart chunking, and markdown formatting of scanned PDFs

Convert scanned PDFs into searchable text locally using Vision LLMs (olmOCR). 100% private, offline, and free. Features a modern Web UI & CLI.

DocInferX is a fully-local, privacy-focused document intelligence system. It ingests PDFs and images, performs OCR, cleans text, chunks content, embeds it into a vector database, and lets you chat with your documents offline using a lightweight LLM (Phi-2).

Documind is an intelligent desktop assistant that allows users to upload documents (PDF, DOCX, TXT) and ask natural language questions about their contents. It leverages local LLMs via Ollama, integrates advanced OCR for scanned files, and uses semantic indexing with LangChain + FAISS to deliver fast, context-aware answers.

A Python tool to automatically process, categorize, rename, and organize scanned PDF documents using OCR and a local LLM.

This repository features a Python script that automates the organization of PDF documents by leveraging a locally hosted Language Model (LLM). The script efficiently extracts text from PDFs using PyMuPDF, with an OCR fallback for more complex files, and classifies them based on their relevance to biological life sciences. By utilizing a local LLM,

A performant Retrieval-Augmented Generation (RAG) pipeline using Ollama to run local LLMs like gpt-oss:20b, llama3:8b, qwen3:4b, and gemma3:4b. Supports PDFs, JPGs, PNGs via OCR, uses Sentence-Transformers for embeddings and ChromaDB for vector storage. Includes benchmarking tools for speed, memory, and answer quality across models.

Streamlit PDF OCR app using Qwen3-VL vision models running locally on Apple Silicon via MLX

Local-first OCR → Markdown → RAG toolkit with optional Hugging Face/custom endpoints. Tk UI for viewing, OCR-ing, and chatting with PDFs; Docker/pyproject ready; pluggable OCR/embeddings/LLM.

Automatically extracts, validates, and structures invoice data from images and PDFs using OCR + a local LLM—eliminating manual data entry.

An AI-powered local research paper analyzer. PDF to structured Markdown via Ollama (OCR + LLM) with a built-in Web UI. (基于 Ollama 的本地 AI 论文解读工具。支持 PDF 到结构化 Markdown，集成 OCR 与 LLM 分析，自带 Web 界面。)

Built a fully offline AI Assistant combining voice commands, local LLMs (LLaMA), vision (OCR, image captioning), and system control. Enabled natural voice Q&A from documents/screenshots, app launcher, and PDF search without API usage. Stack: Python, LangChain, LLaMA.cpp, OCR, Whisper, TTS, FAISS

End-to-end pipeline for building a Kazakh history Q&A bot that extracts text from PDFs using OCR, generates training data with GPT-4-mini, and fine-tunes a local LLM (Qwen 2.5) with LoRA on Mac M4

Offline Multimodal RAG System: Ingest, index, and query DOC, PDF, images, and audio using a local LLM. Provides semantic search and context-aware responses entirely offline, supporting OCR, speech-to-text, and vector-based retrieval for versatile, private data access.

PlantDeck is an offline herbal RAG that indexes your PDF books and monographs, extracts text/images with OCR, and answers questions with page-level citations using a local LLM via Ollama. Runs on your machine; no cloud. Field guide only; not medical advice.

A lightweight local AI assistant that lets you chat with your files — PDFs, documents, images, videos, and code — using semantic search, embeddings, OCR, and multimodal LLMs. Optimized to run on modest GPUs (e.g., RTX 3050 4GB) without requiring heavy VRAM like ChatRTX.

Vedavault is an offline AI tutor that learns from PDFs using OCR and RAG, answers questions in multiple languages, supports Whisper-based speech-to-text and TTS voice responses, and generates topic-wise quizzes. Runs fully on local LLMs via Ollama with complete privacy and no internet needed.

This tool automatically detects new scanned PDFs (or images) in a specified folder, extracts text using OCR (Optical Character Recognition), and intelligently renames each file based on its content. A local LLM (Large Language Model) then analyzes the extracted text to create meaningful filenames (including dates) and moves to folders

No description available

OCR de PDF a Markdown con LLM local.

Fully offline invoice PDF to structured JSON extraction using OCR and local LLM

Intelligent Document Processing — Local LLM-powered PDF ingestion, OCR, RAG Q&A, and multimodal chatbot

Aurora OS · Local PDF-to-audiobook engine with OCR, LLM text cleanup and Coqui TTS.

Extract payment data from PDF invoices using a local LLM (Ollama) with automatic scan detection and macOS Vision OCR

Local LLM–powered Streamlit app to extract algebra and calculus math questions from text, PDFs, and scanned documents using OCR.

AI-powered PDF Question Answering system using RAG, FAISS, OCR, and flan-t5-base. Upload PDFs and ask questions with detailed answers using local LLM.

RESTful Spring Boot service that extracts and analyzes text from uploaded images or PDF documents using OCR and a local LLM

Automatically rename PDF files based on their content using a local LLM (Ollama) with automatic scan detection and macOS Vision OCR

A powerful, local-first financial dashboard that uses **AI (LLMs)** and **OCR** to turn messy PDF bank statements into beautiful, interactive insights.

A lightweight pipeline to extract structured invoice fields from PDFs and images using: RapidOCR for robust OCR. Ollama to run a local Qwen Instruct LLM