Privacy-first · Across every AI tool
Live counter + monthly receipt for Claude, ChatGPT, Cursor, Gemini, and Perplexity. Free forever — your prompts never leave your browser.
No card required · Works on all major AI platforms
Available in the Chrome Web StoreReviewed by Google's Chrome Web Store team
Live example
A 10-second taste — type a prompt, watch tokens tally as you go, then see your daily total roll into the dashboard. No signup, no install.
What it does
A live overlay while you type, a unified dashboard for the month, and surgical rewrites that quietly cut your bill. All without ever reading your conversations.
A discreet overlay shows token count and cost in real time as you chat — across every supported platform.
One dashboard for Claude, ChatGPT, Cursor, Gemini, and Perplexity. Daily spend curves, cost-per-message, where it's going.
Auto-rewrites cut tokens, not meaning — the kind of optimization you'd do manually if you had time. One click to apply.
Separate input/output token tracking shows where the money actually goes — usually in the response, not your prompt.
All counting happens locally. We read API metadata only — token counts, model names, timestamps. We never see your conversations.
Sign in once, see yesterday's stats from any device. Data stored with row-level security. Cancel anytime.
Smart Rephrase
Click ✦ on any prompt. Claude Haiku rewrites it tighter — same intent, fewer tokens. Watch a real example below.
Pricing
Free forever for the basics. Pro unlocks AI-powered Rephrase and history that goes beyond a week. Top-up packs let power users go further without paying more per month.
For casual users wanting to see where their tokens go.
For developers, writers, and analysts using AI daily.
Shared dashboard for the whole org. Pro features × every seat.
All plans · 14-day money-back guarantee · cancel any time · billed via Paddle
Chapter 04 · MCP PRO
Connect TokenEyez to Claude Desktop, Cursor, or any MCP-compatible client. Ask plainly. Get answers, in chat, without leaving the conversation.
Works with Claude Desktop · Cursor · Cline · Continue · Zed · any MCP-compatible client
Reading
Field notes on AI token economics, prompt cost discipline, and the small habits that separate teams who spend $400 / month from teams who spend $4,000 for the same output.
Why Nvidia's Jensen Huang said he'd be "deeply alarmed" if a $500k engineer didn't consume $250k in AI tokens — and what it means for the rest of us.
Your AI invoice tells you the total. It doesn't tell you which prompts cost $4 each, or that 30% of your bill is the same five prompts being retried by hand.