MCP Tracker – See the growth of the MCP registry (mcp-tracker.martinalderson.com)

CatchMetrics’ Martin Alderson has launched an MCP Registry Tracker that monitors the Model Context Protocol (MCP) server registry and reveals continued uptake across the ecosystem. The live snapshot shows 1,112 registered servers in total, split between 628 stdio servers (local or process-bound endpoints) and 481 remote servers (network-accessible endpoints), with 21 new entries added in the last day. A growth-over-time view accompanies the counts to highlight momentum and short-term trends. For the AI/ML community this is a clear sign of increasing adoption and ecosystem maturity around MCP — a protocol designed for sharing model context, tool integrations, and interoperable model access. The mix of stdio versus remote servers implies many deployments remain local or tightly coupled to host processes while a sizable and growing remote footprint supports distributed and networked tooling. Practically, the registry improves discovery, versioning, and provenance for model-facing services, while also surfacing needs around authentication, API stability, and trust/vetting of endpoints. Developers and operators can use these metrics for integration planning, capacity estimates, and research into deployment patterns as MCP moves from experimental to production use.