---
title: Usage & Cost
description: "See what your agents actually cost. Official subscription quota for every login, plus token and dollar history mined from Claude Code and Codex logs already on your disk."
updated: "2026-11-14"
---

Agent CLIs tell you almost nothing about spend until you hit a wall. Agentastic's Usage page answers both questions in one place: **how much quota is left right now**, and **what the last week, month, or quarter of agent work actually consumed**.

Open it with **View → Usage** (<kbd>Cmd</kbd>+<kbd>Opt</kbd>+<kbd>U</kbd>). It opens as an editor tab, so you can split it beside a chat and watch the numbers move. Every [space](/docs/features/spaces) has the same page as its **Usage** tab.

The page reads like a document. It opens with one sentence on the limit closest to running out ("Claude is at 71% of its weekly limit, with 2 days left."), or says that nothing is close. A short paragraph under it gives the other limits and what the range came to.

## Limits

Under **Limits** is a row for each *login*, in provider order: **Claude**, **Codex**, and **Grok** today.

Each row shows the provider's own reported rate-limit windows — the 5-hour window, the weekly window, the monthly window, whichever that plan publishes — as a thin meter, the percent used, and when it resets. A meter turns orange from 70% and red from 90%; below that it stays gray. The numbers come from the provider's API using your existing credentials, refreshed every five minutes. The refresh button beside the page's title reads them again now.

When a token needs renewing, the row says so, offers **Sign In…**, and gives you the exact sign-in command to copy.

### More than one account

Most people have more than one login — work and personal, or a team seat and a side project. Agentastic finds them and tracks each separately.

Discovery is deliberately narrow, so nothing goes hunting through your home directory for credentials. It looks at:

- the provider's default directory (`~/.claude`, `~/.codex`, `~/.grok`),
- sibling profile directories you made by convention (`~/.claude-work`, `~/.codex-personal`),
- and directories Agentastic provisioned itself.

**Add account** signs the new profile in from a window in the app. For Claude you choose between a setup token that Agentastic keeps and Claude Code's own sign-in, which runs in a terminal here. For Codex and Grok you get a one-time code to enter on the provider's sign-in page. The resulting profile gets its own config directory, so the two logins never overwrite each other. Rename or remove a profile from the **…** menu that shows when you point at its row. Agents launched against a profile inherit it through the provider's own environment variable — `CLAUDE_CONFIG_DIR`, `CODEX_HOME`, `GROK_HOME`.

### Claude setup tokens

Claude Code keeps its sign-in in a macOS Keychain item that it owns. Copying that sign-in to a cloud VM means asking macOS for it, and macOS can prompt or refuse. A setup token avoids this: run `claude setup-token`, approve it in the browser, then paste the token into **Use a Setup Token…** in the account's **…** menu. Agentastic checks the token with Anthropic and keeps it in its own Keychain item.

- [Cloud agents](/docs/guides/cloud-agents#agent-authentication) on that account sign in with the token.
- On this Mac, Claude Code's own sign-in is used wherever the account has one. The token signs in an account that has none.
- A token lasts about a year and can't be renewed. The account's row warns two weeks before it stops working. **Replace Setup Token…** takes a new one, and **Disconnect Setup Token** removes it.
- Anthropic's usage endpoint doesn't accept setup tokens. For an account signed in only by a token, Agentastic reads the quota from a request for a single output token to the smallest Claude model, no more than once every ten minutes unless you refresh by hand. That request counts against the subscription.

### Choosing which providers appear

Which providers show up follows the per-agent **Show Subscription Usage** toggle in [Settings → Agents](/docs/reference/settings#agents). That same toggle puts the remaining quota on the agent's row in the [Inspector's Agents card](/docs/features/ide#inspector), where you see it without leaving your work.

With none enabled, the page shows every supported provider and tells you where the toggle lives — a blank page reads as broken rather than unconfigured.

## Token history

The rest of the page is built from logs the CLIs already write to your disk: Claude Code's per-project JSONL transcripts and Codex's rollout logs. Nothing is uploaded, and no extra instrumentation is needed — history exists for work you did before you ever installed Agentastic.

Getting this right is harder than summing a column. Claude Code writes a *growing* usage snapshot per content block, and resume, fork, and compaction can rewrite the same API call into several files — so every event carries a global dedup key and only the largest snapshot survives. Codex emits both a running total and a per-turn delta, and re-emits the same delta on UI refreshes. Sum the wrong field and your numbers inflate quadratically. Agentastic reads the right ones.

### What you get

The section's heading names the range ("Last 30 days"). Beside it, pick the measure and the range:

| Control | Options |
|---|---|
| Measure | **Cost** or **Tokens** |
| Range | **7 days**, **30 days**, or **90 days** |
| Legend | Click a provider to hide it from the chart |
| Chart | A bar for each day; click a day to see its numbers in the margin |

The chart stacks the providers in each day's bar and tells them apart by how dark they are, not by color: color on this page is kept for limits that are running out.

**By model**, under the chart, lists cost, share, and tokens per model, so you can see what an expensive default is really costing you.

The margin beside the column (under it, in a narrow window) holds the rest:

| In the margin | Meaning |
|---|---|
| The day you picked | Cost and tokens per provider for that day |
| Processed | Everything the models read and wrote |
| Cached input | Cache reads, and their share of all input-side tokens |
| Uncached input | Tokens billed at full input rates |
| Output | Generated tokens |
| Saved by caching | Dollars prompt caching saved, and that saving as a multiple of what the range cost |
| By project | Cost per project directory, so you can tell which repository is eating the budget |

Saved by caching is usually the surprise. On a heavy agent workload, cache savings routinely run several times the actual bill.

### About the money

Costs are **estimates at published pay-as-you-go list rates**, never read from a bill. The rate table is versioned in the app with a visible freshness stamp and models each provider's long-context tiers — where a request that crosses a context threshold bills the *whole* request at elevated rates, not just the overflow.

If you are on a subscription, treat the dollar figures as "what this would have cost on the API" — which is exactly the number you want when deciding whether a plan is worth it.

### Speed

The corpus is skewed hard toward old, enormous logs: the last week is a fraction of a gigabyte while the full quarter can be tens of gigabytes. So scans deepen in stages — 7 days, then 30, then 90 — publishing after each, and stop at the range you are actually looking at. Usable numbers land in about a second; widening the range is what asks for the deeper read.

Scanning runs on a dedicated background queue, so the page never blocks the app, and previous results stay on screen while a rescan runs.

## Scope

The Usage panel covers **this Mac**. Agents running on [remote hosts](/docs/guides/remote-machines) or [cloud VMs](/docs/guides/cloud-agents) write their logs on those machines and aren't included yet.

## Related

- [Supported Providers](/docs/features/providers) — what each agent can do
- [Settings Reference](/docs/reference/settings#agents) — the Show Subscription Usage toggle
- [IDE Overview](/docs/features/ide#inspector) — quota on the Inspector's Agents card
