AI / LLM · PRO TIER

Khojpro

Khoj is a personal AI assistant that searches your notes, documents, and the web — answering questions in chat with citations from your own data. Self-hosted, with native clients for Obsidian, Emacs, browser, mobile. Khoj indexes Markdown, PDF, images, org-mode, and chats with the content via any LLM.

Install via WHMCS → Visit khoj.dev ↗

🤖 AI / LLM Min 2048 MB RAM Port 42110 (http) Tier pro

// What it is

A closer look.

It's the "personal Perplexity" — your second brain made conversational, on your infrastructure.

// Use cases

What it's for.

Concrete scenarios where teams pick Khoj over the SaaS alternative.

◆

Personal knowledge management

chat with your Obsidian vault, Apple Notes, journal

◈

Document Q&A

ask questions about your PDF library

◇

Research assistant

combine your notes + web search in one query

▣

Daily planning

agent capabilities for calendars, todos, reminders

▦

Multimodal search

find images by text description in your local photo archive

// Who it's for

Built for these teams.

If your team profile matches one of these, Khoj is a strong fit out of the box.

Profile A

Knowledge workers

with growing note libraries (Obsidian, Logseq, Bear)

Profile B

Researchers & students

wanting AI Q&A on their reading list

Profile C

Power users

building a "second brain" with chat-first access

Profile D

Privacy-bound users

who can't push personal notes to ChatGPT

Profile E

Productivity geeks

experimenting with personal AI assistant workflows

// Differentiators

Why teams pick Khoj.

When evaluating self-hosted options for this category, here are the dimensions on which Khoj consistently lands above the alternatives.

✓Native Obsidian / Emacs plugins — chat with notes from inside the editor
✓Multimodal — handles text, images, PDFs in one index
✓Online + offline modes — search local + web in same query
✓Apache 2.0 — fully open, fork freely
✓Agent capabilities — calendar, web search, automation built in
✓Multiple client interfaces — web, mobile, desktop plugins

// Integrations

Connects to.

The stack you'll plug Khoj into — services, protocols, and adjacent apps in the BluixApps catalog.

◇

LLM backends

Ollama (local) or OpenAI / Anthropic / cloud

◈

Note systems

Obsidian vault, org-mode files, Markdown directories

◆

Document formats

PDF, DOCX, plain text, images via OCR

▣

Web search

built-in Google / DuckDuckGo / Bing

▦

Calendar

Google Calendar, iCal sync

▩

Authentication

OAuth (Google, GitHub), magic link

▼

Client plugins

Obsidian, Emacs, web, iOS/Android

// Adoption & deployment

Notable users & community

28k+ GitHub stars
Strong Obsidian community adoption
Active Discord and GitHub discussions
Featured in PKM (personal knowledge management) influencer videos
Backed by Khoj AI with sustainable open-core model

What we ship

Docker compose: Khoj server + persistent index volume
Pinned ghcr.io/khoj-ai/khoj:latest (release-tagged)
HTTPS via Let's Encrypt
Admin user with random password on first boot
Auto-detects Ollama on same VPS for local LLM mode
Webhook endpoint for integration with Obsidian plugin
Backup hook covers index + database

// Tips & operations

Run it properly.

Operational guidance from running this in production — what to do before you scale, what to lock down, what surprises people.

// PERFORMANCE

Pair with Ollama

for fully offline personal AI — no notes leave your network

// SECURITY

Index incrementally

bulk import of large note libraries takes time; let it run

// OPERATIONS

Set rate limits

public-facing instance can be abused; cap queries per user

// RELIABILITY

Persist /data

index + embeddings live here; mount volume from day one

// DEPLOYMENT

Memory usage

indexing large note libraries needs more RAM than runtime; size accordingly

// SCALING

Update notes via API

Khoj watches filesystem; bulk changes need re-index via API

2048

// min ram (MB)

// min disk (GB)

42110

// access port

http

// protocol

pro

// bluixapps tier

pgvector/pgvector:pg15 · ghcr.io/khoj-ai/khoj:latest

// docker image

Project resources

Official sitekhoj.dev ↗

// Alternatives in AI / LLM

Compare with