---
title: "Discussion on \"Claude Code: subscription or API? Work it out from your own logs\" | Hashnode"
description: "Discussion on \"Claude Code: subscription or API? Work it out from your own logs\". Originally published at neverpaytwice.com.\nShort answer: price one month of your real Claude Code usage at API rates and put it next to the plan price. If the API figure is lower than the plan, paying"
url: "https://hashnode.com/posts/claude-code-subscription-or-api-work-it-out-from-your-own-logs/6afc964723bcee5798db1b0e"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Claude Code: subscription or API? Work it out from your own logs](https://neverpaytwice.hashnode.dev/claude-code-subscription-or-api-work-it-out-from-your-own-logs?utm_source=hashnode&utm_medium=feed)

Originally published at neverpaytwice.com. Short answer: price one month of your real Claude Code usage at API rates and put it next to the plan price. If the API figure is lower than the plan, paying

3d ago · 5 min read · neverpaytwice.hashnode.dev  
Read the article

-   [#claude](https://hashnode.com/tag/claude)
-   [#llm](https://hashnode.com/tag/llm)

## Written by

[NeverPayTwice](https://hashnode.com/@neverpaytwice)

@neverpaytwice

  
[View profile](https://hashnode.com/@neverpaytwice)

Free, tested guides and MIT scripts on real AI token usage and API costs.

Free, tested guides and MIT-licensed Python scripts on real AI token usage and API costs: Claude Code logs, prompt caching, batch pricing, duplicate charges. Originals at https://neverpaytwice.com

## Responses 3

Markdown supported · ⌘↵ to post

[Tim Zinin](https://hashnode.com/@timzinin)

[1d ago](https://hashnode.com/posts/claude-code-subscription-or-api-work-it-out-from-your-own-logs/6afc964723bcee5798db1b0e/comment/6afef1c8778dcdfa4abaeee9)

Separating cache reads from writes and counting each response once makes this comparison much more useful. When a log contains a model with no listed price, does the script mark that month's total as incomplete or only list the model separately?

[NeverPayTwice](https://hashnode.com/@neverpaytwice)

[1d ago](https://hashnode.com/posts/claude-code-subscription-or-api-work-it-out-from-your-own-logs/6afc964723bcee5798db1b0e/comment/6afef1c8778dcdfa4abaeee9#reply-6aff1310c2cbe8a0687505f5)

Right now it only lists the model separately. After the monthly table it prints "Skipped (no price)" with each unpriced model and how many responses it had, and those responses are left out of the monthly totals. The month rows themselves are not marked incomplete, and the skipped count is per model across all your logs rather than per month, so a month that used an unpriced model will read low without saying so. That is a fair gap; for now, adding the model's price to the table at the top of the script and running it again gives a complete total.

[Tim Zinin](https://hashnode.com/@timzinin)

[1d ago](https://hashnode.com/posts/claude-code-subscription-or-api-work-it-out-from-your-own-logs/6afc964723bcee5798db1b0e/comment/6afef1c8778dcdfa4abaeee9#reply-6b0012b7a33f96a400bdcbf8)

The missing per-month marker could make a low total look complete. Would a per-month count of skipped responses be enough to flag that gap without adding estimated prices?
