Frequently asked questions

Argus is an anti-detect browser for running many separate online identities at once. It comes as two programs: a launcher, which is the app you sign in to and where you manage profiles, proxies, cookies and extensions; and the browser itself, a purpose-built Chromium that actually opens the sessions.

No. You install the launcher, and it downloads and installs the browser engine for you on first run. There is no second installer to find and no separate update to keep track of.

macOS 13 Ventura and newer on Apple Silicon, and Windows 10 and 11 on x64. There is no Intel Mac build — an Apple Silicon machine is required on macOS.

Yes. The macOS build is signed by SIMNETIQ LTD with the hardened runtime, and it is notarized by Apple and stapled, so it opens normally with no Gatekeeper warning and no right-click workaround. The Windows build is distributed as a standard installer.

No, and this is deliberate. The launcher hands the browser one launch payload — a profile id, a fingerprint, proxy settings, a data directory, extension paths — and nothing else. The browser has no credentials, cannot sign in, and has no knowledge of your other profiles or your organization's data.

Your profile definitions, proxies, cookie sets, folders, status labels and shared extensions sync to your organization so they follow you between machines. The actual browsing session — cache, local storage, and the live cookie jar — stays in that profile's own directory on the machine that ran it.

Yes. Argus takes HTTP and SOCKS5 proxies, including SOCKS5 proxies that require a username and password — which ordinary Chromium cannot use at all. Proxies live in a shared library rather than being retyped per profile, and health checks report the egress IP, country and latency.

Health checks run automatically for new and previously failing proxies, and any proxy can be re-checked on demand. If the proxy assigned to a profile fails its check, the launch is blocked rather than allowed to proceed and expose your real address.

With the email address and password you used to create your account on this site. The desktop app uses email and password sign-in only — there is no Google or social sign-in.

Yes. Members of an organization share one pool of profiles, proxies, cookie sets, folders and extensions. Writes are made one row at a time, so two people working simultaneously cannot overwrite each other's changes.

Extensions you add — from a Chrome Web Store link or a local folder — are shared with your organization and assigned to the profiles you choose. The four bundled built-ins are toggled at the organization level, so switching one off switches it off for everyone's profiles, not just yours.

There is a local automation API built into the desktop app: a HTTP bridge on your own machine with scoped keys you can create and revoke, covering profile and proxy CRUD, fingerprint updates, launching sessions and authoring automations. It can also be set up as an MCP integration so an agent can drive it. Every endpoint is documented on the API reference page. It is a local desktop API — there is no hosted web API today.

In the launcher, under the API tab. The raw key is shown once, at creation — only a hash is stored, so a lost key is replaced rather than recovered. A key can be scoped to specific folders so it only reaches the profiles inside them, and the first request from any new client raises an approve-or-deny prompt in the app before anything is served. Keys are never created on this website, because the API lives on your machine.

Set Argus up as an MCP server from the launcher's Integrations tab. For most clients — Claude Code, Codex, Cursor, Gemini CLI and others — the launcher writes the config file for you and shows you the path first. Once connected, the agent sees Argus as a set of tools: listing profiles, launching sessions, reading pages. The integrations page documents every supported client and the snippet for doing it by hand.

An automation is a tree of steps run against one browser profile, through that profile's proxy, cookies and fingerprint: navigate, click, type, extract, run scripts, ask a model, branch, loop and send a notification when it finishes. It runs in the desktop app on your own machine — there is no hosted runner, so nothing executes while the launcher is closed. The automation page documents every step type and its fields.

Paid plans are bought in USD through Revolut. Create an account, sign in, then choose a plan from the pricing page to go through checkout. Each purchase covers a single 30-day term and does not renew automatically — nothing is charged again unless you buy it again.

There is no recurring charge to cancel, so nothing is needed to stop paying. If you want to drop to the Free plan before your 30 days are up, you can do that from your dashboard at any time. Fees already paid are non-refundable except where required by law.

Each plan sets how many browser profiles the workspace can keep, how many people can share it, and how many automations it can store. Those limits are enforced in the database rather than only in the interface, so they behave the same on every machine you sign in from. See the pricing page for the current tiers.

Sign in and use the support page to open a request, or email us directly. We reply to the address on your account.