# MCP server overview

The [Model Context Protocol](https://modelcontextprotocol.io/docs/getting-started/intro)
(MCP) standardizes how AI applications and agents connect to outside data
sources. MCP servers can be local or remote. Local MCP servers typically run on
your local machine and use the standard input and output streams (stdio) for
communication between services on the same device. Remote MCP servers run on the
service's infrastructure and offer an HTTP endpoint to AI applications for
communication between the AI MCP client and the MCP server.

For more information, see [MCP architecture](https://modelcontextprotocol.io/docs/learn/architecture).

Bigtable supports the following remote servers for usage with MCP:

- **Bigtable servers**:

  - Admin MCP server: Provides tools for managing Bigtable instances and tables.
  - Data MCP server: Provides the `execute_sql` tool that you can use to query Bigtable data using natural language prompts.

  For more information, see [Use the Bigtable MCP server](https://docs.cloud.google.com/bigtable/docs/use-bigtable-mcp)
  and view the detailed [Bigtable MCP tool reference](https://docs.cloud.google.com/bigtable/docs/reference/admin/mcp/bigtable/mcp).
- **Database Insights server**:
  Provides tools for querying and analyzing database performance and system
  metrics.

  - Learn how to [Use the Database Insights MCP server](https://docs.cloud.google.com/bigtable/docs/use-database-insights-mcp).
  - View the detailed [Database Insights MCP tool reference](https://docs.cloud.google.com/bigtable/docs/reference/admin/mcp/databaseinsights/mcp).