A curated list on AI token economics: what tokens cost, where they get wasted, and how to cut the bill. Tools, benchmarks, papers, and copy-paste configs for the token economy of LLMs and coding agents.

OpenSquilla — Token-Efficient AI Agent with same budget, higher intelligence density

MCP server for token-efficient large document analysis via the use of REPL state

Portable project memory across Claude Code, Codex and OpenCode, plus token accounting measured from harness transcripts. Local file I/O, no API calls, no telemetry.

Unofficial MCP server for working with Kagi *without* API access (you'll need to be a customer). Searches and summarizes. Uses Kagi session token for easy authentication. Works well with your agent.

Context engineering for AI agents. ~80% fewer tokens. Fix tool overload. Skills and memory with in-process BM25 and semantic retrieval. Progressive Disclosure. No vector DB.

Free Kagi Search MCP server — access search and summarizer via session tokens, no API key needed. Works with Claude Desktop, Cursor, Windsurf, Hermes and any MCP client.

Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support.

MCP for semantic code search & navigation that reduces token waste

An agentic memory database that cuts session tokens by 82–99%. One portable SQLite file — your agent's memory, anywhere.

Quickly reads webpages and converts to markdown for fast, token efficient web scraping

Code Mode inspired local sandboxed MCP Gateway - collapses N servers x M tools into 2 tools (~1,000 tokens)

Token conscious database MCP server for Postgres, MySQL, SQL Server, MariaDB, SQLite.

🧠 The right skill, one API call. AI agent skills registry with token-efficient skill resolution. 5,000+ skills from 500+ top repos.

MCP server providing token-efficient access to OpenAPI/Swagger specs via MCP Resource Templates for client-side exploration.

Agentic-MCP, Progressive MCP client with three-layer lazy loading. Validates AgentSkills.io pattern for efficient token usage. Use MCP without pre-install & wasting full-loading

Claude Code session log viewer for JSONL files in ~/.claude/projects. Browse conversations, tool calls, tokens, and live tail sessions on desktop, web, and TUI.

Extract domain knowledge from codebases to reduce LLM token consumption by 20x and time in agentic search by 10x — gathers and makes concepts, naming conventions, and vocabulary queryable via MCP.

Save 30% token costs when using Claude Code, Codex, OpenCode for free - with open source, local semantic search. Works for small and large codebases and monorepos! Enterprise-ready and fully compliant via Ollama and SQLite-vec.

Own your LLM's web search: a local search->fetch->rank pipeline that replaces hosted web-search tools. Measured: matches hosted accuracy at 66% lower cost and up to 88% fewer tokens, plus a precision-tuned semantic caching with query-dependant TTL that no API offers.

Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.

Helping coding agents never make mistakes working with public or private libraries without wasting the context window.

A third-party Perplexity MCP/REST API implementation that leverages Pro account Pool to provide unlimited quota for ask and deep search capabilities.

The open source, no-code MCP Server for AI-Native API Access