How to Check Your AI Coding Token Usage
Every AI coding tool already logs your usage locally. The trick is reading all of those logs at once.
Quick answer
npx whoburnedmore in your terminal. It reads the usage logs Claude Code, Codex, Gemini CLI and 12+ other tools already keep on your machine, then totals your tokens and estimated cost — across every tool — in a few seconds. Only the totals ever leave your computer. 🔥If you run more than one AI coding agent, your usage is scattered across a handful of hidden log folders. Claude Code keeps its own. Codex keeps its own. Gemini CLI, Copilot, Cursor and the rest each keep theirs. There's no single screen that says “here's everything you burned this month.” This guide shows the fastest way to get that number — and the per-tool deep dives if you only care about one.
Your logs
already on disk
whoburnedmore
reads + tallies
Your totals
tokens + cost
Leaderboard
optional, shareable
How do I see my total token usage?
You don't need to install anything or sign in. One command reads what's already there:
$ npx whoburnedmore↳ scanning local usage logs… TOOL TOKENS COST claude code 18.4M $61.20 codex 9.1M $24.80 gemini cli 4.7M $6.10 ───────────────────────────── total 32.2M $92.10
How is the cost calculated?
Tokens aren't dollars, so whoburnedmore converts them using each model's public per-token price. The total cost is simply the sum, over every tool, of that tool's tokens multiplied by its model price:
Estimated, not billed
On a flat-rate plan (Claude Pro/Max, a ChatGPT plan) you don't pay per token — so this figure is what your usage wouldcost at API rates. It's the honest way to tell whether your subscription is a bargain or a rip-off.Check usage tool by tool
Want the details for a single agent? Each guide below covers that tool's built-in commands, where it stores logs, and the gotchas:
- 1
Claude Code
The built-in/usageshows your window; for full history see how to check Claude Code token usage. - 2
Codex CLI
/statusonly shows the current 5-hour window — see your full Codex usage and cost. - 3
Gemini CLI
No built-in spend view at all — read your Gemini CLI usage from its logs.
Which approach should I use?
The built-in commands are fine for a quick glance at one tool. A reader like whoburnedmore wins the moment you run more than one agent, or want history and a single number:
| Approach | Multi-tool | History | Cost estimate | Shareable |
|---|---|---|---|---|
| Built-in /usage, /status | — | — | partial | — |
| A token counter site | — | — | guess | — |
| whoburnedmore | 12+ tools | leaderboard |
tools supported
code or prompts sent
command to rank
Does this send my code anywhere?
No. whoburnedmore reads aggregate usage numbers — token counts and timestamps — and never your prompts, code, or file names. You can prove it with npx whoburnedmore --dry-run to see exactly what would be submitted, or run npx whoburnedmore --local to keep everything on your machine and just open the dashboard in your browser. 🛡️
Related guides
How to Check Claude Code Token Usage
See your Claude Code tokens by day, model, and project — and how the built-in /usage compares.
AI Coding Cost: Claude Code vs Codex vs Gemini
A real cost comparison of the big three — measured from your logs, not marketing.
The AI Coding Token Leaderboard: Who Burned Most?
The fun part: rank your token burn against every other developer on the board.