What does MCP stand for?
Model Context Protocol. It is an open standard for connecting AI assistants to tools and data, introduced by Anthropic in November 2024 and now supported by many AI clients.
Is MCP only for Claude?
No. MCP is an open protocol. Many clients support it, including Claude, ChatGPT, Cursor, VS Code agents and several agent frameworks. Check your specific client for its current level of support.
Does MCP replace our API?
No. An MCP server usually calls your existing API. It adds a thin layer designed for AI models: fewer, clearer tools plus permissions, limits and logging.
How long does it take to build an MCP server?
A focused server for one system with up to ten tools usually takes one to two weeks. Multi-system servers with OAuth and audit logging take three to six weeks.
How much does a custom MCP server cost?
Typical market prices are $1,500 to $5,000 for one system and $5,000 to $15,000 for production multi-system servers. Our prices start at $1,900 and $5,900, fixed in advance.
Can the AI delete or change our data?
Only if you expose tools that do so. A well-built server is read-only by default, exposes narrow write tools only where needed and asks a person to confirm risky actions.
Where does our data go?
The MCP server runs where you choose: your cloud, your network or managed hosting. The AI client receives only what a tool returns for a specific question.
What is the difference between local and remote MCP servers?
Local servers run on each user's computer over stdio. Remote servers run at one URL over Streamable HTTP, usually with OAuth sign-in, which makes permissions and audit easier for teams.
Do we need a developer to maintain it?
Someone must own it. APIs and AI clients change, so servers need occasional updates. You can do it in-house or use a care plan like ours at $390 per month.
What if our vendor already has an official MCP server?
Use it if it covers your fields, permissions and audit needs. Teams build custom servers for custom objects, guarded writes, combined systems or data residency.