private beta, invite-only. no invite? join the waitlist →

built to empower the
next generation of research. 

Modular Investigation,
Notes & Exploration platform for Researchers, by researchers,
Versatile & Adaptive
every idea, connected
papers, notes, and citations in one living graph
your research, your machine, your way
local-first, your own ai, fully adaptable
ask, and it gets to work
a real research assistant, not a chat box

no invite yet? join the waitlist.

leave your name and email, and we send invite codes in small batches as the beta opens up.

we only use this to send you an invite. no newsletter, no third parties.

a look inside minerva

Real views from the workspace. Collect, read, connect, and find your footing, all in one place.

minerva knowledge graph, hundreds of linked papers and notes beside an ai chat panel
01

Every idea, connected

Papers, notes, and citations come together in one living graph. Follow any thread across your whole library and watch how your thinking fits together.

minerva paper library, a searchable list of papers with a pdf preview open alongside
02

Your whole library, in one place

Search every paper, note, and tag from a single list, with the full PDF a click away. Nothing scattered across folders and browser tabs.

minerva pdf reader, a paper with highlighted text and a notes panel of extracted equations
03

Read and annotate in place

Open the PDF next to your notes, highlight as you read, and pull equations and figures straight into your own writing.

minerva research assistant importing a paper from an arxiv link, showing the import tool result and a reply that offers the next step
04

Ask, and it gets to work

Hand the assistant a link and it imports the paper for you, figures and all. It searches your library and the open web, reads alongside you, and runs on a cloud model or one kept fully on your machine.

01 What exactly is Minerva?

One workspace for your whole research journey. Collect papers, read and annotate them, connect every idea into a single searchable graph, write up what you find, and bring in AI whenever it helps, instead of juggling a dozen disconnected tools.

02 Who is it for?

Every researcher, in every discipline. The humanities, social sciences, life sciences, and computational fields all read, connect, and write, and Minerva is built around that shared work rather than one narrow niche. It fits a first-year PhD student and a tenured professor alike.

03 How is it different from a reference manager or a notes app?

Those keep your papers, your notes, and your writing in separate places. Minerva connects them. Every paper, citation, and note feeds one graph you can search and follow in any direction, and you read, annotate, and write without leaving the workspace.

  • Paste from arXiv, DOI, or any URL, figures included
  • PDF and notes open side by side
  • Jump from any citation to its source
  • Spot references that aren't connected yet
04 What does a typical day in Minerva look like?

Paste a link and Minerva pulls the paper, its figures, and its references in seconds. Read it with your notes beside it, highlight by dragging, and everything you capture joins one searchable graph. When it is time to write, your sources and a full editor sit side by side, with Markdown, math, links between notes, and a compile-to-PDF step built in. The pages and tools you reach for open in the same window, so a stray browser tab never pulls you away.

05 Do I have to use AI, and which models can I use?

No, you never have to. AI stays out of your way until you ask for it. When you do, it works as a full research assistant rather than a chat box bolted on the side. It can search across your library and the open web, find and import new papers, read what you have open, and help you draft, always with the model you choose. You bring your own provider, and your keys stay on your machine, used only to reach the model you picked.

  • Claude, OpenAI, and OpenRouter, or a local model fully offline
  • Searches your library and the open web
  • Finds, imports, and reads papers on request
  • Custom agents, reusable skills, and a memory you control
06 What happens to my papers, notes, and API keys?

The real work happens in the Minerva desktop app on your own machine, not in someone else's cloud. Your AI keys stay local, and paper discovery runs on OpenAlex, the free and open scholarly index. Your library and your notes stay yours.

07 What does it cost?

Minerva is free during the private beta. It is also built to run on free, open services, OpenAlex for finding papers and OpenRouter's free model tier for AI, so you can work end to end without paying for anything.

08 Will it still fit me as my research grows?

Yes. Minerva shapes itself around how you work and keeps adapting as your needs change, from your first PhD paper to a professional career. It is one workspace you are not meant to outgrow.

  • A saved layout for every project
  • Rebind every keyboard shortcut
  • Themes and typography you choose
  • Reusable templates for notes and papers
09 Who is building Minerva, and how do I get in?

Minerva is built by researchers at the University of Manchester, with backgrounds across leading technology and research labs, who got tired of stitching their work together from a dozen tools. It is by researchers, for researchers. It is in invite-only private beta right now. If you do not have a code yet, the waitlist is how you get one, and we send invites in small batches as access opens up.

no invite? join the waitlist →

your research, your way

Every project gets its own canvas. Dock the panels the work needs, a reader, your notes, the graph, an AI assistant, a terminal, and a built-in browser, then arrange them into a workflow of your own. Minerva remembers the layout for next time.