See how much of every AI subscription you have left, account by account.

No. Brimline only reads the sign-in that Claude Code keeps. It never renews a token, never uses the refresh token, and never writes to the Keychain or to your Claude Code folders. That is precisely why an expired account stays expired until you open Claude Code with it.

Only inside the usage request to the provider that issued them (Anthropic for Claude, OpenAI for ChatGPT, xAI for Grok, Cursor for Cursor), over HTTPS. They are never sent anywhere else, never written to the cache, to a log, to the command's output or to the MCP server's answers. There is no Brimline server. For Gemini, Brimline sends nothing itself: it launches Google's command, which does what it does with Google.

Access tokens are short-lived. Claude Code renews its own when you use it, and Brimline deliberately does not. Open Claude Code with that account and use it once (run /login if it asks); the next reading is up to date. Meanwhile Brimline shows the last known numbers with their age.

Because the provider reports none for that seat: some team seats have a five hour session limit but no weekly limit across all models. Brimline shows exactly that, “No weekly limit across all models”, and still shows a weekly limit for one model if there is one.

No. Brimline is a Mac app only: macOS 14 or later, Apple silicon or Intel. It lives in the macOS menu bar and reads what the AI tools keep on the Mac; there is no Windows or Linux version.

A Developer ID signature and Apple's notarization go through the paid Apple Developer Program, which Brimline does not use. The app is signed ad hoc, and its source code is public: you can read every line of what runs, or build it yourself. A disk image downloaded in a browser therefore needs one approval the first time (System Settings, Privacy & Security, Open Anyway): how.

Yes. No price, no paid tier, no account to create.

Yes. The PolyForm Shield 1.0.0 license lets anyone use, modify and share Brimline, at work too. What it forbids is offering a product that competes with Brimline. Your employer's own rules on tools and sign-ins still apply.

Five: Claude (every Claude Code account on your Mac), ChatGPT through Codex, Gemini through Google's Antigravity command, Grok through the Grok CLI, and Cursor. Each one is read through the official tool that holds the sign-in, and every one except Claude is off until you turn it on in Settings. Claude has been read on real accounts since version 0.1.0; each new provider was tried on a real account on 27 October 2026, only on the plan named in Link your accounts; nothing is claimed for a plan that was not tried.

That the real answer of a real account was read, that its shape matches the test data, that ten reads in a row gave the same numbers, that nothing was written or renewed at the provider (except that Google's program may renew its own sign-in for Gemini) and that turning the provider off stopped every read. It is said per provider and per plan, never for a plan that was not tried. Two comparisons with the provider's own usage page are still pending (ChatGPT through Codex, and Cursor), Gemini's percentage was seen at 100 % remaining only, and a free Grok account gets no limit from xAI, which Brimline says in words.

Because xAI returns none for that account: on the free account that was tried, the answer has no credit limit. Brimline says exactly that, “xAI returns no credit limit for this account.”, and never shows a 0 %. What a paid Grok account answers was not tried, and nothing is claimed for it.

We do not know, and we do not claim it is. Anthropic's documentation (Authentication and credential use) says OAuth sign-ins are meant for Claude Code and Anthropic's own apps, and that third-party developers may not collect, store or intermediate Claude.ai credentials or session tokens. Brimline reads your own token on your own Mac, read only, never writes, renews or copies it, sends it only to Anthropic to ask for your own usage, and announces itself as Brimline. It is not an official or endorsed product. The same page says Anthropic may enforce these restrictions without prior notice: the risk, if any, is on the account Brimline reads. Read the terms of your plan and judge for yourself.

No. Brimline is an independent project, with no link to Anthropic, OpenAI, Google, xAI or Anysphere (Cursor).

The usage endpoints are internal and undocumented, so they may change. When an answer no longer looks as expected, Brimline marks the accounts “format changed” or unavailable instead of showing a wrong number, and a new version adapts to the change.

Its source code is public, but it is not open source in the OSI sense: the license is PolyForm Shield 1.0.0, which forbids competing products. Everything else is allowed: reading, using, modifying, sharing.