Introduction

Omnilib is an AI-powered research workspace that brings together everything you need for research, development, and creative work. Instead of switching between a text editor, PDF reader, terminal, notebook environment, and chat window, you work in one cohesive place — with AI woven directly into every part of the experience.

Why Omnilib?

Unified Workspace

Traditional workflows fragment your attention across applications. Omnilib consolidates them:

• Multiple panels — View and edit files side by side without alt-tabbing
• Tabbed interface — Keep dozens of files open and switch between them instantly
• Drag and drop — Rearrange the layout to match how you think

AI That Understands Your Project

Omnilib's AI knows your project. It reads your files, understands your bibliography, remembers what you've been working on, and can take action — not just answer questions:

• AI Chat — Ask questions, get explanations, and analyze documents in natural language
• Agent Mode — Let the AI read files, write code, run commands, and edit documents autonomously
• Plan Mode — Have the AI design an implementation plan before executing multi-step tasks
• Sub-agents — Spawn parallel exploration agents for complex research tasks
• Project Memory — A bi-temporal knowledge graph that persists context across sessions

Built for Research

Omnilib was designed with researchers, academics, and knowledge workers in mind:

• PDF annotations — Highlight, draw, and add notes directly on papers
• Jupyter notebooks — Run Python code with real kernel support
• LaTeX compilation — Write and compile LaTeX with live PDF preview
• Bibliography management — Import, organize, and cite references
• Semantic search — Find relevant files and papers using natural language

Key Features

File Management

• File tree navigation with color-coded icons by file type
• Global search with fuzzy matching (Cmd+F (Windows: Ctrl+F))
• Multi-panel layout with draggable tabs

Document Viewing

• PDF Viewer — Full-featured viewer with annotations, highlighting, and search
• Image Viewer — Zoom, pan, and pinch support
• Markdown Editor — Write with live rendered preview

Code and Development

• Code Editor — Syntax highlighting for 50+ languages
• Terminal — Native shell with full PTY support
• Notebook Support — Execute Jupyter notebooks with Python kernels

AI Integration

• Multiple Model Providers — OpenRouter (free and paid), GitHub Copilot, or local models via Ollama
• Agent Behaviors — Five built-in behaviors (Generalist, Coder, Writer, Teacher, Researcher) plus custom ones
• Voice Input — Push-to-talk dictation in desktop apps
• Floating Assistant — Draggable AI chat that stays accessible in any context
• MCP Server — Expose Omnilib tools to external AI agents like Claude Code and Cursor

Collaboration

• Real-time co-editing with presence indicators (online collaborator avatars)
• Share buttons for documents and workspaces
• Comments and annotations

Community

• Built-in community feed inside the app
• Access roadmap, feedback board, and changelog without leaving Omnilib

Git Source Control

• Visual Git panel in the sidebar
• Stage, commit, diff, and push without leaving the workspace

Billing

• Manage your subscription and token usage directly in Settings

Subscription Tiers

Additional usage is available as credit packs. You can upgrade, downgrade, or cancel at any time from Settings > Billing.

What Can You Do?

For Researchers

• Read and annotate research papers alongside your notes
• Write papers in LaTeX with live PDF preview and bibliography support
• Run data analysis in Jupyter notebooks
• Use AI to summarize papers, explain methods, or draft sections
• Build and query a project memory that understands your domain

For Developers

• Edit code with syntax highlighting, completion, and AI-powered refactoring
• Run commands in the integrated terminal
• Use Agent mode to write tests, refactor functions, or debug issues
• Connect external AI agents via the MCP server

For Writers

• Write in Markdown with live preview
• Compile LaTeX documents with Tectonic
• Export to PDF
• Use AI to help with drafts, editing, and structure

Platform Support

Omnilib runs on:

• macOS — Full desktop experience with native features including voice input and local model support
• Windows — Full desktop functionality on Windows 10 and 11
• Web — Access your workspace in any modern browser at web.omnilib.ai

Note: Some features — local models via Ollama, voice input, and the MCP server — are available on the desktop app only.

Related

• Quick Start — Set up and start using Omnilib
• Interface Overview — Learn about the workspace layout
• Project Templates — Start faster with templates
• AI Chat — Begin working with AI assistance