AI & MCP Tools
Tools for building with large language models: work out what an API workload will cost, see how many tokens a prompt uses, and connect AI clients to tools through the Model Context Protocol (MCP).
MCP is an open protocol that lets AI applications such as chat assistants and code editors use outside tools and data. Each client keeps its list of servers in its own config format; the MCP tools here write those files and find what is wrong with them.
Tools in this category
LLM Cost Calculator
Estimate what an AI feature costs per request, day, month and year on OpenAI, Claude, Gemini and more, with cached and batch prices and their sources.
AI Token Counter
Count tokens in your browser: exact for GPT models with OpenAI's o200k_base and cl100k_base tokenizers, estimated for Claude, Gemini, DeepSeek and Mistral.
MCP Config Generator
Write or convert MCP server configs for Claude Desktop, Claude Code, Cursor, VS Code, Devin and Codex, and add servers to the file you already have.
MCP Config & Schema Validator
Check MCP configs for Claude, Cursor, VS Code and Codex, and MCP server tool schemas against the spec, with the line of each problem and a fix where safe.
Which tool do I need?
| Your task | Tool |
|---|---|
| Compare what different models would cost for your workload | LLM Cost Calculator |
| Count the tokens in a prompt or a document | AI Token Counter |
| Write the config that adds an MCP server to your AI client | MCP Config Generator |
| Find out why an MCP server does not load, or check a tool's schema | MCP Config & Schema Validator |