Categories / MCP & Agent Integration Infrastructure
2 · First VoyageMCP & Agent Integration Infrastructure
Gateways, registries, and runtime infrastructure that sit between AI agents and the tools/data they call via the Model Context Protocol (MCP) -- providing discovery, authentication, policy enforcement, and observability for agent-to-tool traffic.
Last updated October 27, 2026
Our read
Anthropic released the Model Context Protocol (MCP) in 2024. Within about a year, at least five venture-backed startups were building the gateways, registries and runtimes that sit between AI agents and the tools they call. Before MCP, teams connected agents to tools through API gateways like Kong and Apigee or integration platforms like Zapier and Workato. None of them supported MCP's tool discovery, so teams wrote glue code that broke whenever an agent or server changed. We count 60+ active companies and no unicorns, and the largest round is Obot AI's $35M seed in 2025. Kong, Microsoft, IBM and Docker have all put out their own MCP tools as well.
Open question for foundersEvery company in this category builds on a spec it doesn't write. If your product depends on a standard you don't control, what's still yours after the next release?
By the numbers
How they talk about AI
Of the 9 pitching agents with public pricing, 3 still price per seat.
How we classify
We read each company's homepage headline and subheadline against a fixed word list; the most explicit match wins. Not mentioned: no AI language at all. Implied: automated, intelligent, smart or autopilot, without saying AI. AI: AI, machine learning, LLM or generative. AI agents: agents, agentic, autonomous or digital workers.This reads positioning, not product: a company with a strong AI product counts as "not mentioned" if its homepage doesn't say so. Homepages are re-read every month.
In their own words
- Arcade.dev
“Govern every action for every AI agent”
Arcade is the actions runtime between your agents and every system they reach. One control point to enforce, execute, and govern every action, every time. Teams keep shipping. Security stays in control.
- Composio
“Give your AI secure access to 1,500+ apps in minutes”
Make your AI finish what you ask. Composio gives it the right tools, access to your accounts, and skills that already work, so it gets it right the first time.
- IBM
“ContextForge”
An open source registry and proxy that federates MCP, A2A, and REST/gRPC APIs with centralized governance, discovery, and observability. Optimizes Agent & Tool calling, and supports plugins.
- Kong
“AI Connectivity for the agentic era”
Secure, manage, speed up, monetize, and cost-control every LLM, MCP, event, and API request.
- Lunar.dev
“Beyond an MCP Gateway: Control every AI interaction across your enterprise”
A unified, enterprise-grade gateway to secure and govern all AI applications and agents
- Manufact
“The cloud for production MCP”
Get on ChatGPT and Claude plugin stores, understand how agents use them and test every release across clients and harnesses.
- Microsoft
“Microsoft Entra Agent ID”
Manage, govern, and protect agent identities and their access to resources—just as you do for employees—with identity, lifecycle, and access controls, now available in Microsoft Agent 365.
- Mintive
“Give every team agents they can trust”
Every agent gets the models, tools, and data it needs to work alongside your team. Mintive connects them through one platform that governs access by role and records every action.
- TrueFoundry
“The Enterprise AI Gateway for LLMs, MCPs & Agents”
On-prem, VPC, hybrid, or public cloud
- Zapier
“Welcome to automation in the agentic era”
Start building in your favorite AI tool, then run and govern trusted workflows on Zapier.
Stage history
- First Voyagerecorded October 12, 2026
Companies we track
Leaders
Challengers
FAQ
What is MCP & Agent Integration Infrastructure?
Gateways, registries, and runtime infrastructure that sit between AI agents and the tools/data they call via the Model Context Protocol (MCP) -- providing discovery, authentication, policy enforcement, and observability for agent-to-tool traffic.
How mature is the MCP & Agent Integration Infrastructure category?
We place it in stage 2, First Voyage. A few pioneers are sailing the same route. The category has a name but no clear leader — education is still your primary job.
How do MCP & Agent Integration Infrastructure companies usually price?
There's no single dominant model among the 10 tracked companies with a public pricing model: 4 flat rate, 3 hybrid, 3 per seat.
When did the MCP & Agent Integration Infrastructure category start?
The first company we count in the category dates to 2024.
How do MCP & Agent Integration Infrastructure companies talk about AI?
Of the 10 we track, by the language in their homepage headline and subheadline as of November 1, 2026: AI agents 9, AI 1, Implied 0, Not mentioned 0. Of the 9 pitching AI agents with public pricing, 3 price per seat.
Which MCP & Agent Integration Infrastructure companies does First to Market track?
Arcade.dev, Manufact, IBM, Kong, Microsoft, Zapier, Mintive, TrueFoundry, Composio, Lunar.dev.
Building in MCP & Agent Integration Infrastructure?
Ten questions, about 90 seconds. Benchmark your category against our library of 238 companies across 27 categories, or find the categories that have already been where you are.