MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
Agentic-first token counting and text chunking service for LLM context management. Count tokens, chunk text, truncate to limits, RAG-style splitting, estimate API costs. Plain text API, MCP-enabled, single Go binary, zero dependencies, no database.
-
csv-token-counter-mcpgithub.com/yourusername/csv-token-counter-mcpupdated this year
MCP server to count tokens in CSV and Excel files
pypi MIT License -
@justmpm/token-counterupdated this year
MCP Server that estimates token counts for files and directories
npm MIT -
io.github.rccola990-cloud/mcp-token-optimizergithub.com/rccola990-cloud/mcp-token-optimizerno update data
Cut LLM token costs: count tokens, estimate cost, slim prompts, and pick the cheapest capable model.
npm -
Precise token counting and context efficiency analysis for MCP servers
npm MIT -
@justmpm/calculatorupdated this year
MCP Server with calculation tools: token counting, math expressions, and more
npm MIT -
SigRankno update data
SigRank — ranks AI coding operators by token cascade efficiency (Yield = Cache Reads × Output / Input²). Tools for yield calculation, leaderboard queries, operator lookups, local token usage pulls, and signed token-count submission to the public board. Privacy-first: runs locally, only four token counts leave the machine (ed25519-signed). Install: npx sigrank
smithery -
@kunal_dutt/token-lensactive
Count and cut the tokens your prompts, chats and agents use, adapted to the model and task. MCP server and self-hosted gateway. Collects no data.
npm MIT -
An MCP server for analyzing codebases with accurate token counting
npm MIT -
Token counts and LLM cost checks for prompts in your repo: a CI linter, pull-request cost diffs, and an MCP server. Runs locally; prompts are never uploaded.
npm UNLICENSED -
mnemox-mcpgithub.com/ajaygh99/mnemox-extensionupdated this year
Mnemox MCP server — score prompts, count tokens, save/get memories. Works with Claude Code, Cowork, Cursor.
npm MIT -
mcptokensgithub.com/dondai1234/mcptokensupdated this year
375 tokens in your agent's harness. One MCP tool that counts the tool-definition cost of any other MCP server.
pypi MIT -
llm-cost-estimator-mcpgithub.com/waqarulwahab/llm-cost-estimatorupdated this year
MCP server for LLM token counting and multi-model cost estimation (GPT-4o, Claude, Gemini, and 25+ models). Local-first, no API key.
npm MIT -
MCP server for counting tokens using OpenAI's tiktoken library
MIT ★ 2 -
glyph-compressgithub.com/Neolambo/glyph-compressupdated this year
Cut the tokens your IDE-to-LLM coding session is billed for. Stops re-sending files the model already has, places provider cache breakpoints where they cover the request, and compresses what needs sending — measured against real tokenizers, not character counts. CLI, VS Code, MCP server, proxy.
npm AGPL-3.0 ★ 4 -
tokenslimgithub.com/himanshudhawale/tokenslimupdated this year
Shrink the token cost of context you feed to LLMs — count tokens, estimate cost across models, and slim files (with a CI budget guard) before you send them.
MIT -
MCP server for PICO-8 development. Token counting, code analysis, validation, minification, and cart comparison via shrinko8. Includes documentation resources. Compatible with any MCP client.
★ 10 -
A ready-to-clone Rust MCP server for Claude via crimson-crab — ask_claude, chat, count_tokens, list_models, get_model
Apache-2.0 -
markdowninthemiddlegithub.com/rickcrawford/markdowninthemiddleupdated this year
A local proxy that converts websites and APIs to clean Markdown. Convert HTML pages, JSON APIs, and dynamic sites. Get token counts for LLM budgeting.
MIT ★ 8 -
📚 Any ebook → clean token-counted Markdown chapters. Let Claude read your bookshelf over MCP — and catch it misquoting your books. Deterministic RAG: no embeddings, no network, no API keys.
pypi MIT ★ 2 -
mcp-toolkitgithub.com/PranayMahendrakar/mcp-toolkitupdated this year
The basics AI keeps fumbling - real current time, exact math, word counts, regex actually tested, exact diff, token estimator, JSON/YAML validator, JWT decoder. Free MCP plugin for Claude and ChatGPT, zero extra credits.
MIT -
A Model Context Protocol server with machine learning engineering tools: BPE token counting and cost estimates, BM25 search, paired bootstrap and McNemar evaluation comparison, JSON Schema validation, and prompt-injection and PII scanning.
MIT -
Turn a codebase, Git repo or folder into LLM-ready files: exact token counts, secret redaction, repo maps, split parts and reviewed apply-back of LLM replies. Desktop app (Windows, macOS, Linux), CLI and MCP server.
MIT -
mcp-character-toolsgithub.com/Aaryan-Kapoor/mcp-character-toolsupdated this year
The last thing you need for your LLM to work with individual characters or count the number of r's in a word. This is an MCP server providing 14+ comprehensive (and pretty) character and text analysis tools to help LLMs work with individual characters - something they struggle with due to tokenization.
MIT ★ 3 -
Web pages and PDFs to clean, LLM-ready Markdown as a remote MCP tool (Claude, Cursor, VS Code) and via the Apify API. Token counts, small crawls, sitemaps, RAG chunks, respects robots.txt. Python and Node examples.
MIT -
See where your OpenAI Codex tokens, dollars and plan limits go: every response counted once, live 5-hour and weekly limits from your own logs, quota forensics, a web dashboard, menu bar and tray, an MCP server and a share card. Python, no dependencies, no login.
pypi MIT -
Track what every AI coding agent costs and see the limit coming. Per-session, per-model cost for Claude Code, Codex, Cursor and 37 more tools from your provider's own token counts, plus limit gauges, a dangerous-command guard and dashboard terminals. Free, Apache-2.0, runs on your machine.
go NOASSERTION ★ 60 -
token-lens-mcpgithub.com/mdfifty50-boop/token-lens-mcpupdated this year
MCP server for context window optimization — audit token budgets, compress schemas, optimize tool loading
npm -
token-proctorgithub.com/navintkr/token-proctorupdated this year
VS Code extension + MCP server that validates prompts, routes to the cheapest LLM, and projects token × turn cost before the call. Cuts Copilot premium-request burn on agent loops.
MIT ★ 3 -
Any web page to LLM-ready Markdown: one binary, MCP server, browser extension
MIT -
只回数字、不回正文的文本统计 MCP 服务器。单文件实现,只依赖 @modelcontextprotocol/sdk, 按路径统计行数/汉字数/词数/字节数等口径,结果中绝不包含文件正文。
MIT