How to Check Amazon Q Developer Usage and Limits

Amazon Q Developer is AWS's coding agent that lives both inside your IDE and in the terminal through the q command. Like most assistant products in 2026 it runs on a freemium model: the free tier hands you a modest monthly allotment of agentic requests— the higher-effort, multi-step actions where Q plans, edits files, and runs tools on your behalf — and the paid Pro tier lifts that ceiling. Because the free allotment is small, “how many Amazon Q requests do I have left this month?” is one of the most common things developers go looking for. The honest answer is that AWS itself is the only place that number lives, so this guide points you there first, then shows where a tool like whoburnedmore fits for the rest of your stack.

How do I check my Amazon Q usage?

Amazon Q meters agentic requests against your AWS identity, so the authoritative reading is inside AWS — not on your local disk. There are two practical paths, the console and the CLI:

1. 1

   Open the Amazon Q usage view in the AWS console

   Sign in to the AWS console, open Amazon Q Developer, and find the usage or subscription panel. It shows agentic requests consumed against your current monthly allotment for the account or builder ID you are signed in as.
2. 2

   Confirm which identity you are billed under

   Q usage attaches to an AWS Builder ID or an IAM Identity Center user. If you switch between a personal Builder ID and a work SSO login, each has its own separate allotment — a frequent source of “why is my count different?” confusion.
3. 3

   Ask the q CLI from the terminal

   With the agent installed, the q command can report your plan and remaining capacity without opening a browser. Run q --help to find the subcommand your installed version exposes for account and usage details.

$ q whoami↳ builder-id: dev@example.com↳ plan: Free   AMAZON Q DEVELOPER — current cycle  ─────────────────────────────────────────────  Agentic requests:    used / monthly allotment  Resets:              start of next monthly cycle  ↳ exact numbers come from AWS, not your local logs

What is the Amazon Q free-tier limit?

The free tier of Amazon Q Developer includes a small monthly cap on agentic requests, with editor completions and lighter interactions generally not drawing from that same bucket. AWS adjusts the exact figure over time and varies it by surface, so rather than quote a number that may already be stale, check the current allowance on the official Amazon Q Developer pricing page. The shape of the offer, however, is stable: free gives you a taste, Pro removes the squeeze.

1. 1

   Free tier — a small monthly agentic-request allotment

   Enough to evaluate Q and handle the occasional multi-step task. Heavy agent use — a long refactor where Q plans, edits, and re-runs tests — can spend the allotment surprisingly quickly. 💸
2. 2

   Pro tier — a raised limit for daily driving

   Aimed at developers who lean on the agent every day. The cap moves up substantially and the friction of watching a counter mostly disappears.
3. 3

   Check the live figure before you plan around it

   Because AWS revises both the free allotment and Pro pricing, treat any specific number you read elsewhere as a hint, not a contract — the pricing page is canonical. 📊

How do I track my other AI coding agents?

This is where whoburnedmore comes in — honestly and specifically. whoburnedmore does not read Amazon Q. Q keeps its own meter on AWS, and there is no local log for whoburnedmore to parse. What whoburnedmore does do is read the on-disk session logs that other terminal agents leave behind — Claude Code, Codex CLI, Gemini CLI, opencode and friends — and roll them into one token-and-cost table. If Amazon Q is one tool among several you run, this is how you see the others side by side:

$ npx whoburnedmore↳ scanning local agent session logs…↳ Amazon Q not included — Q meters itself on AWS   TERMINAL AGENTS — last 30 days  ─────────────────────────────────────────────  claude code         21.4M     $72.10  codex cli            8.7M     $23.40  gemini cli           5.2M      $6.80  opencode             2.1M      $3.90  ─────────────────────────────────────────  total               37.4M    $106.20   Amazon Q usage → check the AWS console / q CLI

Notice the table is explicit about the gap: it lists the agents whoburnedmore can read and tells you to look elsewhere for Q. That separation is the whole point — you get a real cost picture for your token-metered agents, and you keep using AWS's own view for your request-metered one.

Request-metered vs token-metered: pick the right tool

The mental model that keeps this straight is simple. Amazon Q is request-metered: AWS counts discrete agentic actions, so AWS is where you check it. The agents whoburnedmore supports are token-metered: they record input and output tokens in local logs, so a local reader can sum them. Trying to force one tool to report on the other's metric only produces misleading numbers.

How is the combined agent cost worked out?

For the agents it can read, whoburnedmore turns raw token counts into an API-equivalent dollar figure by pricing each model's input and output tokens at its public rate and summing across tools. Amazon Q is deliberately absent from this sum because Q's cost is a flat subscription plus a request allotment, not a per-token charge you could reconstruct from a log:

Read that as: for every tool tin your local logs, multiply its input tokens by that model's input rate, add its output tokens times the output rate, and total it up. The result is what the same work would have cost on a metered API — a useful sanity check on whether each subscription earns its keep. For Amazon Q, the equivalent question is answered differently: did you stay inside the free agentic-request allotment, or is Pro worth it for your cadence?

What whoburnedmore does and does not cover here

To be unambiguous: whoburnedmore does not connect to your AWS account, does not read your Amazon Q agentic-request count, and cannot show your Q limit. That data lives on AWS and requires AWS authentication. whoburnedmore only reads the local token logs of the other terminal agents on your machine, and only ever sends numeric aggregates — never prompts, code, or file contents. Run npx whoburnedmore --dry-run to see exactly what would be read and submitted before anything leaves your machine. 🛡️