Everything you need to know about Librarian. Can't find your answer? Get in touch.
Does Librarian send my documents to the cloud?
No. Librarian runs entirely on your local machine. Your documents are indexed and stored in a local vector database on your own hardware. Nothing is uploaded, transmitted, or shared with any external server. In Offline mode, the application has zero network access.
What is Agentic RAG?
Agentic RAG (Retrieval-Augmented Generation) is an advanced approach where the AI agent performs multiple retrieval steps to answer your question. Unlike simple RAG that retrieves once and responds, Agentic RAG can break complex questions into sub-queries, gather evidence from different parts of your documents, and synthesize a more complete answer — reducing missed details and hallucinations.
Can I choose which local model Librarian uses?
Yes. Librarian works with any model available through Ollama. You can switch between models depending on your task — smaller models (8B parameters) for fast document Q&A, or larger models (30B+) for complex coding questions. The choice is yours, and you can change models at any time.
Are small models enough for coding help?
For basic code questions and documentation lookup, 8B models work well. For serious coding assistance — debugging, refactoring, or cross-repository analysis — we recommend 30B+ parameter models. Librarian's hardware requirements page has specific recommendations by use case.
What file formats does Librarian support?
Librarian supports PDF, Excel (XLSX/XLS), CSV, plain text, Markdown, and common code files across 20+ programming languages including Python, JavaScript, TypeScript, C#, Java, Rust, Go, and more. Your documents are parsed and indexed locally for fast retrieval.
Can multiple people use Librarian on the same network?
Yes, with the Family plan. One machine runs Librarian as a LAN server, and up to 5 approved devices on the same local network can connect to it. Access is controlled through built-in device trust management — you decide which devices are allowed.
What are the minimum hardware requirements?
For basic document search with small models: 8GB RAM, a modern quad-core CPU, and 10GB free disk space. For larger models and coding assistance: 16–32GB RAM and a dedicated GPU (NVIDIA recommended) significantly improves performance. See our full hardware requirements on the Product page.
How do I get my license key after purchase?
After completing your purchase through Stripe, you'll receive your license key immediately on the confirmation page and via email. The key activates your chosen plan and is tied to your account for easy management.
Is there a free trial?
Yes. The Local Trial plan gives you 30 days of full access at no cost. You can index documents, run queries, and evaluate Librarian on your own hardware before deciding to subscribe.
What operating systems are supported?
Librarian currently supports Windows (10 and later). Linux and macOS support are in active development and coming soon. The application requires Ollama to be installed for local model execution.
What happens if I don't use Librarian for a long time? Will I lose access?
No. Your subscription stays active as long as payments continue. Librarian keeps a 30-day offline grace period so you can use the app without an internet connection. If more than 30 days pass without the app contacting our servers, you'll be asked to connect to the internet briefly on next launch. Once connected, your license re-validates instantly and everything resumes — no data is lost, no re-installation needed.
Still have questions?
Check out the product overview for more detail, or start your free trial and see for yourself.
We use cookies to improve your experience. Essential cookies are required for the site to function. You can choose to enable analytics and marketing cookies. Privacy Policy
Cookie Preferences
Essential CookiesAlways Active
Required for the website to function. These cannot be disabled. They include session management, security features, and basic site functionality.
Analytics Cookies
Help us understand how visitors interact with the website by collecting anonymous usage data such as page views, referral sources, and device information.
Marketing Cookies
Used to measure advertising effectiveness and deliver relevant content. These may be set by third-party partners such as Meta (Facebook/Instagram).