Token burn dashboard

AI usage by day, source, and work driver.

This starter reads one normalized file, labels exact and estimated values, and keeps the question pointed at the work: what should the computer do next?

Total burn14228.0Mselected window
Peak day1219.8M2026-06-27
7d average619.9Mmoving average
Active days31rows in view

Daily burn

Heatmap

Log color scale so quiet days and spikes can share one surface.

Weekly trend

Log-scaled trend

A smooth read on whether usage is getting sharper or merely larger.

Source split

Exact beside estimated

The source labels are part of the dashboard, not a footnote.

exact0Codex / 0%
exact14228.0MClaude Code / 100%
estimated0Claude chat / 0%
estimated0ChatGPT / 0%

Drivers

What is burning tokens

Keep driver labels boring and consistent: shipping, research, review, video, admin.

shipping100%

Scale equivalents

Make the number human

Approximate comparisons are useful only when the math stays visible.

Approx. words10671.0Mtokens x 0.75 words
Reading time711399h250 words per minute
Novel equivalents118566.590k words per novel

Peak day

shipping

shipping across 10 projects

Date2026-06-27local bucket
Burn1219.8Mall sources

Moving-average table

Last 30 days

Exact and estimated columns stay separate.

DateTotal7d avgCodex exactClaude Code exactCallsClaude chat est.ChatGPT est.Driver
2026-07-091080.0M619.9M01080.0M583900shipping
2026-07-08920.9M581.1M0920.9M420400shipping
2026-07-06883.9M579.3M0883.9M452700shipping
2026-07-05421.7M467.9M0421.7M308500shipping
2026-07-04140.4M511.6M0140.4M70000shipping
2026-07-0362.1M665.8M062.1M21200shipping
2026-07-02830.6M702.2M0830.6M332800shipping
2026-07-01808.4M653.5M0808.4M204700shipping
2026-06-30907.9M616.3M0907.9M309900shipping
2026-06-29104.0M541.5M0104.0M44100shipping
2026-06-28728.1M593.0M0728.1M469600shipping
2026-06-271219.8M582.9M01219.8M598100shipping
2026-06-26316.9M427.5M0316.9M189800shipping
2026-06-25489.6M404.8M0489.6M246500shipping
2026-06-24547.5M431.4M0547.5M297100shipping
2026-06-23384.7M464.5M0384.7M197100shipping
2026-06-22464.1M487.6M0464.1M267100shipping
2026-06-21657.9M447.5M0657.9M356900shipping
2026-06-18131.6M373.7M0131.6M62900shipping
2026-06-17157.9M358.1M0157.9M33600shipping
2026-06-16676.2M409.9M0676.2M304100shipping
2026-06-15779.0M321.5M0779.0M447200shipping
2026-06-14546.3M215.4M0546.3M338700shipping
2026-06-13183.8M138.2M0183.8M58100shipping
2026-06-12141.3M112.1M0141.3M112600shipping
2026-06-1122.2M107.2M022.2M17600shipping
2026-06-10520.5M124.2M0520.5M308200shipping
2026-06-0957.4M25.2M057.4M53700shipping
2026-06-0836.2M14.4M036.2M29300shipping
2026-06-026.3M3.6M06.3M6500shipping

Replace data/daily-burn.sample.json with your normalized daily rows. Keep raw exports and private project names out of anything you deploy or share.