Fahad Manzur
Updated May 2026 · 4 min read
Anthropic’s Model Context Protocol has crossed 97 million installs in just 16 months, becoming the universal standard for connecting AI agents to external tools. Here is what MCP is, why it exploded, and what it means for developers and businesses.
Monthly SDK installs (March 2026)
Public MCP servers
To reach 97M installs
What is Model Context Protocol (MCP)?
Model Context Protocol is an open standard created by Anthropic that defines how AI agents connect to external tools, data sources, and APIs. Think of it as USB-C for AI: before MCP, developers had to build a separate custom connector for every combination of AI model and external tool. With MCP, you build a server once and any MCP-compatible AI agent, whether Claude, ChatGPT, Gemini, or others, can use it immediately.
MCP was launched in November 2024. By March 2026 it had reached 97 million monthly SDK downloads, the fastest adoption curve for any AI infrastructure standard in history. Kubernetes, now considered foundational cloud infrastructure, took nearly four years to reach comparable scale. MCP did it in 16 months.
Why did MCP win so fast?
Three factors drove the explosion. First, Anthropic open-sourced MCP immediately under the MIT licence with no fees, no restrictions, and no vendor lock-in. Any company could adopt it without asking permission. Second, the protocol itself is simple: it uses JSON-RPC over standard transports that any language and any runtime can support. Third, Anthropic seeded the ecosystem early with reference implementations for common tools like GitHub, Slack, Postgres, and Google Drive, giving developers a working library to build on before competitors had reason to create alternatives.
Once OpenAI adopted MCP in early 2025 and Google followed in early 2026, the standards war was effectively over. When every major AI lab ships MCP-compatible tooling, choosing anything else means fragmentation.
Who governs MCP now?
In December 2025, Anthropic donated MCP to the Linux Foundation, establishing it under the Agentic AI Foundation (AAIF). Co-founders include Anthropic, Block, and OpenAI. Platinum members include AWS, Google, Microsoft, Cloudflare, and Bloomberg. This neutral governance removed the last hesitation enterprises had about adopting MCP: the risk of a single vendor controlling the specification. MCP now sits alongside Kubernetes, Linux, and PyTorch as infrastructure owned by the industry, not by any single company.
What tools and platforms support MCP today?
On the AI side: Claude, ChatGPT, Google Gemini, Cursor, Windsurf, VS Code, JetBrains AI Assistant, and Microsoft Copilot all support MCP as clients. On the tool side: Google Drive, Slack, Notion, GitHub, Jira, Salesforce, and Postgres all have MCP server implementations. Gartner forecasts that by end of 2026, 75% of API gateway vendors and 50% of integration platform vendors will include MCP support.
What does MCP mean for businesses and developers?
For developers, MCP eliminates the N times M integration nightmare. Five AI models connecting to ten tools previously required fifty custom integrations. With MCP it requires ten MCP servers, usable by all five models simultaneously. For businesses, MCP means AI agents can now connect to your CRM, documents, calendars, and databases without months of custom engineering. Early production deployments show measurable results: one Twilio test reported task success rates rising from 92% to 100% and compute costs dropping up to 30% after switching to MCP-based integrations.
Frequently Asked Questions
What does MCP stand for? +
MCP stands for Model Context Protocol. It is an open standard created by Anthropic that allows AI agents to connect to external tools, APIs, and data sources through a single universal interface rather than requiring a custom connector for each integration.
Does ChatGPT support MCP? +
Yes. OpenAI adopted MCP in early 2025, integrating it across the ChatGPT desktop app and its agent frameworks. This was a decisive moment that confirmed MCP as the industry standard, not just an Anthropic-specific protocol.
Is MCP free to use? +
Yes. MCP was released under the MIT licence from day one and is now governed by the Linux Foundation’s Agentic AI Foundation. There are no licensing fees, and no single vendor controls the standard.
What tools can I connect to AI agents using MCP? +
As of April 2026, MCP server implementations exist for Google Drive, Slack, Notion, GitHub, Jira, Salesforce, Postgres, and thousands more. There are over 10,000 public MCP servers available, covering developer tools, enterprise platforms, and data infrastructure.
How is MCP different from regular API integration? +
Traditional API integrations are model-specific: a connector built for OpenAI will not work with Claude or Gemini without rebuilding it. MCP is a protocol layer above function calling, meaning one MCP server works with any AI agent that supports the standard, regardless of which model powers it.
Who owns and controls MCP now? +
MCP is now governed by the Agentic AI Foundation, a directed fund under the Linux Foundation. It was co-founded by Anthropic, Block, and OpenAI, with AWS, Google, Microsoft, Cloudflare, and Bloomberg as platinum members. No single company controls it.
What is the business impact of MCP adoption? +
Businesses using MCP report faster AI agent deployment, reduced integration engineering costs, and measurably better performance. One production test showed task success rates rising to 100% and compute costs dropping up to 30%. Gartner predicts that by end of 2026, 75% of API gateway vendors will include MCP support, making it effectively unavoidable for enterprise AI workflows.
