Claude Code Token Optimizer is a VS Code extension that, according to its Marketplace listing, analyzes local Claude Code session transcripts to show token-usage patterns and possible areas to review. Its cost and recoverable-savings figures are estimates—not independently verified savings or a promise that your bill will fall.
What the VS Code extension analyzes
The Marketplace listing says the extension reads Claude Code session transcript files stored under ~/.claude/projects/. It presents analysis of token use, cache hits and misses, prompt length, context compaction, and tool or search behavior. The listing says analysis runs locally and data is not sent to a service; that is the publisher’s description, not an independent privacy audit.
![Claude AI for Beginners Bible: [5 in 1] The Ultimate Guide to Automate Your Work, Save Hours Every Week, and Use AI for Real-World Results](https://m.media-amazon.com/images/I/415+fSJacsL._SL160_.jpg)
As an Amazon Associate I earn from qualifying purchases.
Its listed findings and views include:
- Context compactions and high context pressure
- Broad or unscoped searches and oversized command output
- Large fixed prompt overhead, model choice, and fast-mode costs
- Filters, a savings dashboard, trends, regression alerts, per-session drill-down, and a status-bar context indicator
These are capabilities claimed by the Marketplace listing, not independently confirmed here. The extension is described as analyzing existing session data; the available listing does not establish that it changes prompts, limits tool output, or otherwise alters Claude Code’s behavior.
How to find and install it
The listing recommends installing through VS Code’s Extensions view or its Visual Studio Marketplace page. It also documents building a VSIX from source and installing a VSIX through the editor or command line. The listing names ccOptimizer.claudeHome as a setting for overriding the Claude home directory.
#1 Best Overall
Before relying on an analysis, check that the extension can access the Claude home directory and project transcripts on your machine. If you use a non-default Claude home directory, consult the listing’s current setting instructions.
How to interpret cost and savings estimates
The listing says its cost figures use first-party API rates because Claude Code’s local stats report costUSD: 0. For subscription users, it characterizes the resulting figures as API-equivalent costs. Treat them as an estimate of usage translated into API pricing, not as a statement of what a subscription user actually paid.
Rank #2
The extension also estimates recoverable waste. That is not the same as money saved: an estimate identifies possible inefficiency, while realized savings would require a change in workflow and a measured comparison afterward. The listing itself cautions that recoverable waste can be a small fraction of spend when cache reuse is already high. No independent test or realized savings figure is established here, so claims that the extension saves a particular amount are not supported.
Model names and rates can change. The listing points to src/pricing.ts as the maintained rate table; check its current values before interpreting any model-specific cost output. No fixed price or rate comparison is included here.
Don’t confuse it with the separate plugin-plus-VSIX setup
There is a separately documented project called Token Optimizer by alexgreensh. Its Claude Code (VS Code) installation guide describes a plugin that writes data and a companion VSIX that displays it; the VSIX requires the plugin first. That is a distinct integration, not a prerequisite for the Marketplace extension covered above.
Other similarly named Claude Code tools also appear in search results, including a plugin described as supporting multiple coding environments. Similar names do not establish shared code, features, data handling, or installation requirements. Check the project identity and its own setup instructions before installing or comparing tools.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What to verify before trusting a token-saving claim
To judge whether a dashboard represents a useful opportunity rather than a guaranteed result, separate what it observes from what it estimates:
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →- Input: Confirm which local transcripts or other data the tool reads.
- Effect: Determine whether it only analyzes existing sessions or also changes prompts, tool output, or configuration.
- Setup: Check whether it needs a separate plugin, hooks, or other prerequisites.
- Cost method: Find out whether totals reflect actual charges or are calculated from API rates.
- Validation: Compare usage under a consistent workload before and after applying a suggested change; an estimated opportunity alone does not prove a realized saving.
- Data handling: Read the project’s statements about where data is processed and stored. A local-processing claim is useful context, but not an independent security assessment.
The Marketplace listing supports describing Claude Code Token Optimizer as a local-transcript analytics extension with estimated cost and waste views. It does not establish an independently measured reduction in token use or spending.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
