Khoj — Your AI Second Brain

What it is

Khoj is an open-source personal AI application for chat, search, and knowledge management. It connects to your documents, notes, and files to provide AI-powered search and conversation grounded in your personal knowledge base.

Khoj supports multiple LLM providers (OpenAI, Anthropic, local models), integrates with Obsidian and other note-taking tools, and can be accessed via web, WhatsApp, or custom agents. It aims to be your AI second brain.

How it saves time or tokens

Khoj indexes your personal documents and notes, so you can ask questions about your own knowledge base without manually searching through files. The AI retrieves relevant context from your documents before generating answers, reducing hallucination.

By running locally or self-hosted, Khoj keeps your personal data private and avoids per-query costs for document retrieval operations.

Additionally, the project's well-structured documentation and active community mean developers spend less time troubleshooting integration issues. When AI coding assistants generate code for this tool, they can reference established patterns from the documentation, producing correct implementations with fewer iterations and lower token costs.

How to use

1. Install Khoj:

pip install khoj
khoj --anonymous-mode

1. Open the web UI at http://localhost:42110 and configure your data sources (Obsidian vault, markdown files, PDFs, org files).

1. Chat with your documents:

You: What were my notes about the Q1 product roadmap?
Khoj: Based on your notes in planning/q1-roadmap.md, the key priorities were...

1. Use the Obsidian plugin for in-editor AI search and chat.

Example

# Khoj API for programmatic access
import requests

response = requests.get(
    'http://localhost:42110/api/chat',
    params={'q': 'What meetings do I have this week?'},
    headers={'Authorization': 'Bearer your-api-key'}
)
print(response.json()['response'])

Related on TokRepo

• AI Tools for Knowledge Graph — Knowledge management and graph tools
• AI Tools for Research — AI-powered research and search tools

Common pitfalls

• Not re-indexing after adding new documents. Khoj needs to index your files to search them. Set up automatic re-indexing or manually trigger it after adding content.
• Using Khoj with too many large files without chunking configuration. Adjust chunk size and overlap settings for optimal retrieval quality on large document collections.
• Expecting Khoj to replace a full RAG pipeline. Khoj is designed for personal knowledge management. For enterprise document search, consider dedicated RAG solutions.
• Failing to review community discussions and changelogs before upgrading. Breaking changes in major versions can disrupt existing workflows. Pin versions in production and test upgrades in staging first.