A market terminal for people who do their own research.

The ordinary way to research a stock is a chart in one tab, fundamentals in another, a discounted cash flow in a spreadsheet, a screener somewhere else and a broker app for the price. Each tool is competent on its own and none of them talk to each other, so a good part of the work is reconciling figures that should agree, rather than thinking about the business.

Suros is built by Suros Systems, and it is the workspace we wanted for our own research: change the symbol in one panel and every open panel follows, edit a growth assumption and the sensitivity grid moves with it.

No exchange, broker, issuer or asset manager pays for placement in Suros, and there is no order flow to sell. It is paid for by the people who use it, and it answers to them.

The equities terminal covers listed companies and crypto: charts, valuation models, a screener, a portfolio with risk and stress tests, options and futures, paper trading with a prop challenge, and a backtester with trade analysis. The screener scans more than twenty thousand listings from the terminal's own security master, crypto pairs sit alongside them, and search also opens listings on these exchanges.

• New York
• Nasdaq
• London
• Frankfurt
• Paris
• Zurich
• Milan
• Madrid
• Stockholm
• Oslo
• Copenhagen
• Helsinki
• Toronto
• Tokyo
• Hong Kong
• Shanghai
• Seoul
• Mumbai
• São Paulo
• Sydney

Market capitalisation is converted to a single currency before anything is ranked, so a Tokyo bank and a New York bank can be compared directly. Prices stay in the currency they trade in, and a portfolio is valued in the base currency you choose.

The commodity terminal has a world map of facilities and shipping, forward curves, spreads, hedging and futures paper trading. Both terminals share one workspace, and the same data is open to your own tools through a REST API, a Python client and an MCP server for AI assistants.

Vantage sits beside the equities terminal: a workspace where a model of your choice writes code, runs it in your browser on the same market data, and calls the terminal's own backtest, valuation and options engines. It answers with the chart and the code behind it, and changes nothing until you apply what it proposes.

For a tour of each window, read how it works. For worked examples on public data, read the research notes.

1. Every number is market data, or computed from it.

   A price, a ratio, a valuation and a backtest are either fetched from a source or calculated from what was fetched. When a figure cannot be fetched, the terminal shows a dash or says the figure is unavailable. It never fills the gap with an estimate, a sample or a number from somewhere else.

2. An error looks like an error.

   A feed that fails says it failed, and the part of the terminal that depended on it reports itself unavailable until the feed is back. An empty result and a broken one are never allowed to look the same. The status page checks the server and the terminal when you open it and keeps ninety days of daily history.

3. The decision stays yours.

   The models are frameworks driven by your assumptions, not answers. A valuation comes with its sensitivity grid, so you can see how far the answer moves when one input does, and a backtest is a simulation over the past, labelled as one.

4. We say what we do not have.

   There is no SOC 2 report, no ISO 27001 certification, no independent penetration test and no 24-hour on-call rota yet. The security page lists these gaps itself, because a page that stays quiet about them implies something it should not.

The terminal opens once you sign in. Your watchlists, portfolios, paper accounts, valuation assumptions and layouts are stored against your account so a second device sees them, and your browser keeps a working copy. The account copy is held in a managed database in the EU, under a key derived from your account email, and read back only for a request that has signed in as you. Deleting the account deletes it, with your alerts and notification subscriptions, and there is no soft-delete window.

We do not sell personal data or use it for advertising. The strategy builder and the filings question box send the prompt and the document excerpts it concerns to a model provider, without your email or your holdings. Vantage, the research assistant, sends more: your message, the chat so far, the names of your project files and whatever its tools read for the answer, which can be a file's contents, market data, and, in a chat you connect to Paper trading or a prop challenge, that account's positions, orders and balances. On the built-in model this goes through our metered gateway to the model provider; with a key of your own it goes from your browser straight to the provider you chose and never through us, and the key stays in your browser. Vantage projects and chats are kept in your browser, not on our servers. We do not keep any prompt or answer, and nothing you type is used by us to train a model. Market data is fetched by our server, which sends the provider the ticker and nothing that identifies you. The exceptions are crypto prices, which your browser reads from the exchange directly, and the optional Data Link, which fetches prices through your own browser.

Who hosts and processes what, where, and the safeguard for each transfer outside the EEA are listed in the privacy policy. Suros is subject to the GDPR as it applies in Norway, and the supervisory authority is Datatilsynet.

The terminal ran as a closed alpha on real market data from June to August 2026, for a handful of users who signed up. The Excel add-in shipped in October, and the early access beta is planned for December, with places released in cohorts from the waitlist. After that, the order of work is set by what beta members report. The roadmap carries the detail, and says plainly which parts have no date.

Questions, data requests, security reports and Community reports each have their own address, and every one is read by a person. They are all on the contact page.