Enterprise DNA
M MCP Servers Developer low

ooples/token-optimizer-mcp

by Various

Intelligent token optimization for Claude Code - achieving 95%+ token reduction through caching, compression, and smart tool intelligence

O

MCP

ooples/token-optimizer-mcp

Added 1 June 2026

#ai #caching #claude #compression #llm #mcp #mcp-server #token-optimization

Overview

This tool reduces token consumption in Claude Code through caching, compression, and smart tool intelligence. It is implemented in TypeScript as an MCP (Model Context Protocol) server and claims up to 95% token reduction. The approach combines local caching of repeated context with compression strategies to minimize usage while preserving essential information.

Best for

Best for
Claude Code users who want to reduce token usage and costs without losing core functionality

Use cases

  • Cutting token costs during extended Claude Code sessions
  • Keeping context windows manageable when working with large codebases
  • Speeding up Claude Code by reusing cached outputs for frequent queries

How to use

Install

npm install -g @ooples/token-optimizer-mcp

Tools exposed

  • optimize_text
  • get_cached
  • compress_text
  • decompress_text
  • count_tokens
  • analyze_optimization
  • get_cache_stats
  • clear_cache
  • smart_read
  • smart_write
  • smart_edit
  • smart_grep
  • smart_glob
  • smart_diff
  • smart_branch
  • smart_log
  • smart_merge
  • smart_status
  • smart_api_fetch
  • smart_cache_api

Tested with

Claude Desktop, Claude Code, Cursor, Windsurf, Cline, VS Code

Notes

This tool reduces token consumption in Claude Code through caching, compression, and smart tool intelligence. It is implemented in TypeScript as an MCP (Model Context Protocol) server and claims up to 95% token reduction. The approach combines local caching of repeated context with compression strategies to minimize usage while preserving essential information.

403 stars on GitHub. Last updated 2026-05-23. Licensed MIT.

Use cases

  • Cutting token costs during extended Claude Code sessions
  • Keeping context windows manageable when working with large codebases
  • Speeding up Claude Code by reusing cached outputs for frequent queries

Pros

  • High claimed token reduction (95%+) can lower costs significantly
  • Combines multiple optimization techniques (caching, compression, tool intelligence)
  • Open source with a strong community (403 stars on GitHub)

Cons

  • Token reduction may compromise output accuracy or completeness in some scenarios
  • Only works with Claude Code through the MCP protocol, not universal
  • Requires configuration and may need tuning for different workloads

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.

Pros

  • High claimed token reduction (95%+) can lower costs significantly
  • Combines multiple optimization techniques (caching, compression, tool intelligence)
  • Open source with a strong community (403 stars on GitHub)

Cons

  • Token reduction may compromise output accuracy or completeness in some scenarios
  • Only works with Claude Code through the MCP protocol, not universal
  • Requires configuration and may need tuning for different workloads
Free 27-page guide

Get the free Developer’s Field Guide

A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.

Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.

No spam. Unsubscribe any time.

Running a business, not writing the code? See the MCP servers picked for operators, and get your first one wired up with us.

Operator picks