# How I Cut MCP Token Usage by 91% (and Learned a Humbling Lesson About Tokenizers)

> Source: <https://dev.to/mcptokensaver/how-i-cut-mcp-token-usage-by-91-and-learned-a-humbling-lesson-about-tokenizers-5hl1>
> Published: 2026-08-16 02:35:21+00:00

When you add MCP servers to your AI coding agent, each one dumps its full JSON schema into context. 255 tools across all servers = 39,964 tokens. On a 128K context window, that's 31% gone before you type a single character.

I was literally paying for JSON syntax overhead. Every API call included `{"content":[{"type":"text","text":"..."}]}`

— 80 tokens to deliver 6 tokens of data.

I built a CLI that sits between your agent and MCP servers. It does three things:

Instead of full JSON schemas, mcptoon presents tools in a compact pipe-delimited format:

```
# Full JSON (287 tokens per tool):
{"name":"search","description":"Search the web","inputSchema":{"type":"object","properties":{"q":{"type":"string","description":"Query"},"n":{"type":"number"}},"required":["q"]}}

# SLIM format (26 tokens):
search|q:s*|n:n
```

255 tools: 39,964 → 3,511 tokens. **91% saved.** Verified with `tiktoken.get_encoding("cl100k_base")`

.

Schemas live on disk in `~/.mcptoon/config.json`

. Your agent runs `mcptoon manifest --slim`

to see what's available, then `mcptoon call <server> <tool> '{"param":"value"}' --toon`

to execute. Only the compressed output enters context.

Tool results come back as human-readable key-value pairs instead of nested JSON:

```
# JSON result (80 tokens):
{"content":[{"type":"text","text":"{\"name\":\"react\",\"stars\":219000}"}]}

# TOON result (12 tokens):
name: react
stars: 219000
```

My first version replaced `null`

with `∅`

(the empty set symbol). I thought I was being clever. Then someone ran it through tiktoken: `null`

= 1 token, `∅`

= 2 tokens. I was literally increasing token count and calling it optimization.

The HN community called me out on it. Fair enough — I hadn't measured before shipping. Now everything is tiktoken-verified. `true`

stays `true`

. `null`

stays `null`

. No unicode tricks.

At GPT-4o pricing ($5/M tokens):

```
pip install mcptoon
mcptoon add fetch --stdio npx -y @anthropic/mcp-fetch
mcptoon manifest --slim    # see what's available, compact
mcptoon call fetch fetch '{"url":"https://example.com"}' --toon
```

Works with any agent that can run shell commands. One config file for all agents — no more reconfiguring for Claude Code vs Cursor vs OpenCode.

3000 lines of Python, 309 tests, zero dependencies.

GitHub: [https://github.com/activeing123/mcptoon](https://github.com/activeing123/mcptoon)

What's your biggest MCP token waste? How are others handling this?