Library / Overview
Free to use
MANUAL ATLAS

The right manual.
The answer, in context.

Name your component. Ask your question.
Find official documentation and read the evidence.

↗ Original sources◇ No API fees⊞ Stored locally
SEARCH YOUR DOCUMENTS

What are you looking for?

Text search ready

Include a manufacturer and model. We look for official documents, then show original passages and page numbers. First AI search downloads approximately 118 MB to your device.

YOUR DOCUMENTATION

A clearer library.

All in one place. Ready when you need it.

0 documents
▤＋

Your next solution is in a datasheet.

Start with a component and question above.
Or add a PDF you already have.

PDF up to 50 MB · Up to 1000 pages · No account

◉ Stored in this browser. Keep your original files as a backup. Clearing browser data removes the library.

START WITH THE ORIGINAL

Go straight to the manufacturer.

Official documentation, from the source. Paste an official PDF or product URL into the search to retrieve it.

EXTERNAL SOURCES ↗

Automatic retrieval depends on manufacturer access. If a source blocks downloads or needs sign-in, we show the original link instead.

↗

A page number beats a guess.

Read the passage from the document. Check the model variant, document version and full context before using the information.

Find the right passage ↑
BUILD YOUR LIBRARY

Give your document a home.

The PDF stays on your device and its text becomes searchable.

Document

Open original ↗Download PDF ↓
HOW IT WORKS

Free. Local. Source-based.

Manual Atlas is an independent document tool. You do not need an account or paid AI service.

Documents and privacy

PDFs, text and bookmarks are stored in your browser’s IndexedDB on this device. They are not shared with other visitors or synced between devices. Keep your originals as backups.

Free AI search

Semantic search starts automatically with your question and uses a downloaded model to compare your question with document passages. It does not generate technical answers or specifications. Enabling it downloads code and model files from jsDelivr and Hugging Face. These providers receive ordinary connection data such as your IP address; PDFs are not sent.

Technical limitations

Scanned PDFs without text need OCR before searching. Text extraction may be incomplete in diagrams and tables. AI matches are suggestions, not approval of electrical work. Check the original document and applicable requirements.

Finding official manuals

To find documentation, your question or model number is sent to our retrieval server and the manufacturer and model may also be sent to DuckDuckGo to discover official manufacturer websites. Downloaded public PDFs pass through that server. PDFs you upload yourself remain in your browser. Do not include passwords, personal data or confidential information in a discovery question. A model match is not a guarantee that every revision is covered.