Glossary
MCP, the Model Context Protocol, is an open protocol through which AI assistants and agents discover and call tools, read resources and use prompts exposed by a server; an MCP router is an MCP server whose tools make the routing decision, and a router that imports other MCP servers as routing targets.
Assistants in editors and agent frameworks reach the outside world through MCP: a server lists its tools, resources and prompts in a standard shape, and the client calls them over stdio or Streamable HTTP. The protocol is what lets one assistant use a file system, a database and a ticketing system without bespoke integrations for each.
What the protocol does not cover is the model the assistant itself runs on - that is whatever the client was configured with. A router exposed as an MCP server fills the gap: tools named route, estimate, recommend, explain and feedback let the assistant ask which model deserves a task, what it will cost, route it under the workspace's policy and report the outcome - metered and traced like any request, with the workspace key in the server's configuration.
The reverse holds too. A router that imports MCP servers turns every tool, resource and prompt they expose into a routing target, ranked next to models and agents under one policy. Provenance checks on tool parameters and resource limits apply before a tool runs, so an argument that arrived from an untrusted source never reaches a tool unchecked.
Questions people ask
OpenSmartRoute is open source and the free plan keeps the full trace of every decision. Type a request in the playground and read the ranked candidates.
Free plan, no card. Fifteen thousand decisions a month with the full trace.