Ork Meter Plus is free for everyone through October 31, 2026. Then every new account gets 7 days of Plus free.

EDOS
Ork Meter
For teams building with AI coding agents

See what your AI coding agents use. Then make every token count.

Ork Meter tracks your Claude Code and Codex usage in one live dashboard. Ork Meter Plus adds a plugin that trims noisy tool output before the agent reads it.

Free to start. Only usage numbers leave your machine.

How it works

Up and running in minutes.

Install the plugin

Add Ork Meter to Claude Code or Codex. It runs on your computer, next to the agent.

Connect once

Approve the connection in your browser. The plugin then reports model, tokens and time as you work.

See spend, then trim it with Plus

Your dashboard updates live. With Plus, the plugin keeps noise out of the model’s context and records each estimated saving locally.

What you’ll see

Your usage, in plain numbers.

Modeled on two founders running AI coding agent crews for 14 days.

EXAMPLE DATA · ILLUSTRATIVE ONLY, NOT A CUSTOMER’S RESULTS
Tokens used · 14 days
21.08B
Observed usage across Claude Code and Codex: input, output and cache.
Rate-card cost · 14 daysESTIMATE
$49,680
The same tokens at published per-token API rates, labelled as an estimate.
Reported charges · 14 days
$280
Two Claude Max plans at $200 and two Codex plans at $100: $600 a month as the providers bill it, $280 for these 14 days. Kept on their own line, never added to the estimate.
Tokens trimmed by Plus · 14 daysESTIMATE
380M
From the plugin’s local ledger of what it kept out of the model’s context.

Tokens per day

Example: tokens per day over 14 daysIllustrative data for a two-person team running AI coding agents. Weekdays range from 1.42 to 2.32 billion tokens; weekends from 0.35 to 0.56 billion.00.6B1.2B1.8B2.4BDay 1 (Mon): 1.42 billion tokens (example)Day 1Day 2 (Tue): 1.78 billion tokens (example)Day 3 (Wed): 1.89 billion tokens (example)Day 4 (Thu): 1.68 billion tokens (example)Day 5 (Fri): 2.05 billion tokens (example)Day 6 (Sat): 0.46 billion tokens (example)Day 7 (Sun): 0.35 billion tokens (example)Day 8 (Mon): 1.86 billion tokens (example)Day 8Day 9 (Tue): 2.21 billion tokens (example)Day 10 (Wed): 2.12 billion tokens (example)Day 11 (Thu): 1.99 billion tokens (example)Day 12 (Fri): 2.32 billion tokens (example)Day 13 (Sat): 0.56 billion tokens (example)Day 14 (Sun): 0.39 billion tokens (example)Day 14
Show as a table
DayTokens (example)
11.42B
21.78B
31.89B
41.68B
52.05B
60.46B
70.35B
81.86B
92.21B
102.12B
111.99B
122.32B
130.56B
140.39B

By tool

Claude Code17.92B
Codex3.16B

Plus: what the plugin trimmedESTIMATE

OptimizationTimesEst. tokens
Oversized tool output reduced1,840201M
Test noise reduced56686M
Repeated reads reduced82654M
Oversized reads bounded13039M

What the trimming is worth: 380M tokens is about $896 at published per-token rates, or about $5 at these founders’ effective plan rate ($280 over 21.08B tokens). On a flat plan the win is headroom: fewer tokens against the plan’s limits, more work before throttling. From the plugin’s local savings ledger. Estimated at about 4 bytes per token.

Free vs Plus

Ork Meter shows the usage. Plus helps you use less of it.

Plus runs inside the plugin on your machine and keeps the full original of anything it reduces, ready for the agent to open. The figures Plus shows are projections from our own benchmarks; your results will vary and are not guaranteed.

FeatureOrk MeterFreePlusPer seat
See your usage
Usage dashboard for Claude Code and Codex✓✓
Tokens by day, tool and model, including cache use✓✓
Estimates and provider-reported charges kept separate✓✓
CSV export and guided setup✓✓
Use tokens better
Reduce oversized tool output; the full output stays on diskClaude Code and Codex—✓
Filter passing-test noise, keep the failuresClaude Code—✓
Shorten repeated reads of unchanged filesClaude Code—✓
Page huge files instead of loading them wholeClaude Code—✓
Cache-expiry nudge before an expensive resumeClaude Code—✓
Loop guard for repeated identical failuresClaude Code—✓
Local savings ledger, with estimates labelled as estimatesClaude Code—✓
Stay in control
Monthly budget with email alerts at 80%, 100% and 120%—✓
Before-and-after comparison of each optimization in your own data—✓
Pricing

Free to see. A simple per-seat price to save.

Free

Ork Meter

$0

Your AI coding usage in one private dashboard.

  • Claude Code and Codex
  • Daily token usage, including cache use
  • Estimates and reported charges side by side
  • CSV export
Start free
Questions

Frequently asked questions

What data leaves my computer?

Usage metadata: tool and model, a request ID, a timestamp, token counts by category, any charge your provider reports, and your connection’s labels. Your prompts, code and AI responses stay on your machine. See the privacy policy.

Where do the Plus optimizations run?

On your machine, inside the plugin. Anything it trims is saved in full to a local file the agent can open.

How do I see what Plus saves?

The plugin keeps a local ledger of every trim, and Plus compares your usage with and without each optimization in your own data.

How does Plus billing work?

Plus is free for everyone through October 31, 2026, then 7 days free for each new account. You choose when to buy, per seat, monthly or yearly.

Start seeing your AI coding usage today.