TokenLens MCP Server

MCP server that exposes TokenLens pricing data as tools for AI agents (Claude Code, Cursor, Windsurf, Claude Desktop, etc.).

Tools

Related MCP server: LLM Benchmark MCP Server

Setup

1. Install

git clone https://github.com/flikQ/tokenlens-mcp.git
cd tokenlens-mcp
npm install   # or: bun install
npm run build # or: bun run build

2. Connect to your editor

Claude Code (one command):

claude mcp add tokenlens -- node /path/to/tokenlens-mcp/dist/index.js

To always use live pricing data from tokenlens.flikq.dev:

claude mcp add tokenlens -- env TOKENLENS_DATA_URL=https://tokenlens.flikq.dev/api/data node /path/to/tokenlens-mcp/dist/index.js

Cursor · Windsurf · Claude Desktop — add to config JSON:

• Cursor: ~/.cursor/mcp.json

• Windsurf: ~/.codeium/windsurf/mcp_config.json

• Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "tokenlens": {
      "command": "node",
      "args": ["/path/to/tokenlens-mcp/dist/index.js"],
      "env": {
        "TOKENLENS_DATA_URL": "https://tokenlens.flikq.dev/api/data"
      }
    }
  }
}

3. Verify

Restart your editor, then ask your agent:

Use the tokenlens recommend_plan tool. I code 3h/day on claude-sonnet.

Data source

By default the server reads bundled data/*.json files from the repo. To always use the latest live pricing data, set:

TOKENLENS_DATA_URL=https://tokenlens.flikq.dev/api/data

The server fetches all data at startup and caches it in memory. If the URL is unreachable it automatically falls back to the bundled local files.

You can also point at a custom data directory:

TOKENLENS_DATA_DIR=/path/to/custom/data

License

MIT