Om-EIntelligence Dashboard

layout lab: hero 1hero 2hero 3hero 4hero 5hero 6 · you are here

Om-E

transforms your browser into an Agentic AI Knowledge Base.

Windows & macOS · Chrome extension · local-first

· explore ·

One agent. Any model.

Swap the brain — keep the memory.

Google Gemini logoGoogle Gemini
OpenRouter logoOpenRouter
Ollama logoOllama
LM Studio logoLM Studio
Hermes logoHermes
No hallucinations
gate

Powered by Hermes Agent

MIT license · agentic core

Google Gemini logoGoogle Gemini
OpenRouter logoOpenRouter
Ollama logoOllama
LM Studio logoLM Studio
Hermes logoHermes

Use your ChatGPT
subscription to log in.

the plan you have is the engine
no api key

or one OpenRouter key · hundreds of models

What is Om-E

An agentic AI inside your Chrome browserGoogle Chrome

Om-E installs on your PC or Mac and moves into Chrome. Everything you capture — pages, PDFs, decks, transcripts — is turned into clean Markdown and written to a vector store on your own machine: searchable by meaning, permanent, never uploaded.

The browser is its body

A Chrome extension wraps the tabs you already work in.

Your machine is its memory

Everything you keep, vectored locally. The library never uploads.

Hermes by Nous Research

Hermes is its mind

The MIT-licensed agentic core — any model you point it at.

Your work is its purpose

Drafts, verifies, saves back — cited so you can check.

Om-E · the agent with a heart

The intake

Captured once. Known forever.

Word logoWord
Excel logoExcel
PowerPoint logoPowerPoint
PDF logoPDF
CSV logoCSV
Markdown logoMarkdown
Web pages logoWeb pages
YouTube transcripts logoYouTube transcripts
Word logoWord
Excel logoExcel
PowerPoint logoPowerPoint
PDF logoPDF
CSV logoCSV
Markdown logoMarkdown
Web pages logoWeb pages
YouTube transcripts logoYouTube transcripts

Simply drag & drop to import.

Becomes markdown

Every file refined into clean, LLM-ready text.

Vectored locally

Indexed on your Mac — the library never leaves.

Ask — RAG answers

Only the passages that matter reach the model.

Or load it whole

Every word read, cover to cover. Your call.

into the machine

gate

Any model, anywhere

Same chats. Same knowledge. Any model you like.

ChatGPT can't answer as Claude. Claude can't answer as Gemini. Om-E can — because the memory is yours, not the model's. Keep one library and pick the brain per chat: something local for the private work, your own server for the team, a frontier model when the job earns it.

On your machine

Run the model on your own hardware. Nothing leaves the desk, nothing is metered, and it still works on a plane.

lm studio · ollama · offline

On your server

Point Om-E at one box on your network and the whole team shares the same engine — on hardware you own, behind your own door.

one endpoint · your hardware

Over the net

Your ChatGPT subscription logs straight in. One OpenRouter key opens every major lab. Bring what you already pay for.

chatgpt sub · openrouter key

Switch the brain as often as you like — nothing you have built goes with it.

gate

World first

The world's first AI context engineering tool.

01

Stop re-briefing your AI

Capture a page, a PDF, a deck once — it's known forever, in every chat.

02

Answers from your stuff

Grounded in what you've captured, cited so you can check — not confident fiction.

03

Your library stays on your machine

Stored and indexed locally. A turn sends only the slice you chose — run a local model and nothing leaves at all.

04

Swap the brain, keep the memory

OpenAI today, Ollama offline tomorrow. Models are rentals — your context is yours.

Built in

Web search
Deep research
Citations
Verification
Personas
gate

Web search

The agent goes and finds out.

Ask something your library can't answer and it goes out for it: searches, opens the results, reads them, and comes back with what matters — sources attached.

Built in. No search key, no API to wire up.

No key setup. Search ships inside the agent — nothing third-party to configure.

It reads, you don't. It opens the result tabs and does the reading; you get the findings.

Sources attached. Every finding cited, checkable against the page it came from.

the full story →
A frosted ice globe floating in darkness, pulling in glowing documents along violet threads
A vault of dark ice — five vast crystalline screens curved around a viewing platform
gate

The machine

Five moving parts. One workspace.

One screen for each. Keep going.

The machine

Five moving parts.

Chats

Where the work happens — every conversation saved, searchable, resumable.

more →

Projects

Chats grouped into one piece of work — contexts isolated, one shared vector they all feed.

more →

Topics

The unit of memory. Anything worth keeping, saved once and reused forever.

more →

Knowledge Bases

Libraries of topics that search by meaning — your private index, on your machine.

more →

Multi-session

Everything runs at once. Fire a research run in one chat, keep working in another.

more →
gate

Built for trust

Every claim cited. Every answer verifiable.

Answers are built from your captured sources, and every claim carries its citation. Run a verification pass and Om-E checks each claim back against the evidence — verdicts stamped under every one, including the ones that fail.

Citations

Sources travel with the answer — whoever you hand it to can check it themselves.

more →

Verification

Score answers against rubrics you define — the judgement is inspectable, not vibes.

more →

Comparison

Two proposals, two policies, two vendors — where they differ, who wins on which criterion.

more →

Templates

Turn captured research into reports, briefs and memos in your format — citations intact.

more →

Local-first

The library never uploads. And never expires.

Your topics, files and keys live on your machine — the index is local, retrieval is local, and the only thing that ever leaves is the turn you chose to send.

how we handle your data →

Rent the cockpit

One flat price. Your own engine.

Om-E is licensed for a flat term — a month, a quarter, a year. You bring the AI you already pay for; we never meter a token or mark up your usage.

see the terms →

Stop losing knowledge in tabs.

Ask the page. Save the answer. Build your knowledge base.