What is the AI usage leaderboard?

A free public ranking of developers by real AI token usage and spend from Claude Code (captured automatically by MyTokenTracker) plus any API usage members report through the SDK wrappers or events API.

Yes. MyTokenTracker is free forever. Tracking your usage and appearing on the leaderboard cost nothing.

How do I join the leaderboard?

Create an account, install the one-line tracker, and turn on the public leaderboard in Settings. You get a shareable profile with your tokens, cost, streak, favorite model, and a 365-day activity heatmap.

Input, output, and cache tokens, estimated cost, model, provider, and platform for Claude Code and any API usage you report, not prompts or content.

Is my profile public?

Your aggregate usage stats and your GitHub name, handle, and avatar appear publicly on the leaderboard by default. We never show your prompts, content, project names, or file paths. You can turn the leaderboard off anytime in Settings.

Leaderboard

Who is shipping the most AI

The public ranking of real AI usage across every model and platform. Tokens burned, API value, days on a streak. Create an account and you are on the board automatically.

Efficiency Tokens API value

All time Last 30 days

Rank Developer Favorite API value

Kabir Acharya @KabirAcharya claude-opus-5-5 $13.3K 32B tokens 2

pinjashakohen @pinjashakohen claude-sonnet-5 $1.7K 3.3B tokens 3

Kevin Champlin @Kevinchamplin claude-opus-4-8 $1.3K 3B tokens 4 P Paul I Reeder claude-opus-5 $20.55 22.5M tokens 5 S sana braham claude-sonnet-4-6 $20.17 73.8M tokens 6

Jezqualix @Jezqualix claude-haiku-4-5-20251001 $10.63 8.4M tokens

Only people who opted in appear here. Manage your visibility in Settings.

How efficiency works

Plan return is the API value of the usage your plan covered, divided by what the plan cost over the same 30 days. Usage from other providers is not counted, so a Codex habit cannot flatter a Claude plan.

Max value: 10× and up
Power user: 3× and up
Paid off: 1× and up
Under plan: under 1×

Ranked once you have confirmed your plan in Settings and logged 5+ active days in the window. Cache reads are the share of tokens served from cache, the cheapest tokens there are.

Check yours in the terminal

No account needed to see your own usage at API prices.

npx mytokentracker

The free AI usage tracker behind the board

MyTokenTracker tracks your real token usage and cost. Claude Code is captured automatically, and GPT, Gemini and other APIs are tracked through a one-line SDK wrapper or the events API. The leaderboard is the fun part. The data behind it is the point.

→ The AI Cost Index, what it costs to run AI over time
→ Value for money, the lowest cost per million tokens
→ Model prices for 4,489 models
→ Community usage, anonymized and aggregate

Questions

How do I join the leaderboard?: Create an account, install the one-line tracker, and turn on the leaderboard in Settings. You get a shareable profile with your tokens, cost, streak, favorite model, and a 365-day heatmap.
Is it free?: Yes, free forever. No paid plans.
What does it track?: Tokens, cost, model, provider, and platform for Claude Code and any API usage you report. Never your prompts or content.
Is my profile public?: Your usage stats and GitHub profile (name, handle, avatar) appear publicly by default. Never your prompts, project names, or file paths. Turn it off anytime.