Is Claude Max worth it for you?
A flat plan pays for itself when your usage would have cost more at API prices than the plan does. You can measure that instead of guessing: one command gives you the number, and this page turns it into a multiple for each plan.
Step 1: get your number
Run this in any terminal. It reads the logs Claude Code keeps on your machine and prices the last 30 days at API list rates. Copy the total.
npx mytokentracker --days 30Claude Code deletes local transcripts after 30 days by default (Claude Code docs), so 30 days is the window most machines can answer.
Step 2: compare it with each plan
Worked example: $340.00 of usage over 30 days. Replace it with your own number.
| Plan | Monthly price | Plan cost for the period | Multiple |
|---|---|---|---|
| Claude Pro | $20.00 | $20.00 | 17.0x |
| Max 5x | $100.00 | $100.00 | 3.4x |
| Max 20x | $200.00 | $200.00 | 1.7x |
Formula: plan cost = monthly price × days ÷ 30. Multiple = your usage ÷ plan cost. Above 1x, the plan bought more than the same money buys at API prices. Below 1x, paying per token would have been cheaper.
Value is not the same as fitting the limits
The multiple answers "is this plan good value for what I used". It does not answer "will this plan let me work the way I do". Claude plans reset on a rolling five-hour session window, and Max plans also have a weekly limit across all models. Max 5x gives five times the per-session usage of Pro, and Max 20x gives twenty times (Anthropic support).
So in the example above, Pro shows the biggest multiple, but someone with that much usage may hit Pro's limits every afternoon. Read the multiple together with how often you hit a limit. Plan prices are from claude.com/pricing and can change; check there before you switch.
Get this every month without doing the maths
With a free account, npx mytokentracker init syncs daily totals in the background and the dashboard builds a plan-value receipt for 7, 30 or 90 days, with a shareable card. Synced history also stays after Claude Code clears its local logs, so the comparison can cover more than one month.
Frequently asked questions
How do I know what my Claude Code usage is worth?
Run npx mytokentracker in a terminal. It reads the usage logs Claude Code already keeps on your machine and prices every model at API list rates. The total for the last 30 days is the number this calculator needs. Nothing is uploaded and no account is needed.
What formula does the calculator use?
Plan cost for the period is the monthly price times the number of days divided by 30. The multiple is your API-equivalent usage divided by that plan cost. A multiple above 1 means the plan bought more usage than the same money would have bought at API prices.
Does a high multiple mean a cheaper plan would be enough?
No. Value and limits are separate questions. Claude plans have usage limits per five-hour session and per week, and Max 5x and Max 20x give 5 and 20 times the per-session usage of Pro. If you regularly hit limits on Pro, a bigger multiple on Pro does not mean Pro fits your work.
How much do Claude Pro and Max cost?
Claude Pro is $20 a month billed monthly. Max 5x is $100 a month and Max 20x is $200 a month, according to Anthropic's pricing and Max plan pages.
Can I compare against a plan price that is not listed?
Yes. Run npx mytokentracker --plan with any monthly price in dollars, for example --plan 30 for a team seat. The CLI applies the same formula.
Does this work for Codex or other coding agents?
The --plan comparison works for any agent the CLI reads, since it only needs a total and a price. The dashboard's monthly plan-value receipt currently knows Anthropic plans only and lists other agents separately.