How to Use the Amplitude MCP in Pydantic AI

Q: How do I initialize the Amplitude MCP Server in Pydantic AI?

Start with pip install "pydantic-ai-slimmcp". Use the unified MCPToolset("http://...") class with your Vinkius endpoint. Pass this toolset into your Agent constructor. Do not use the deprecated MCPServerHTTP class.

Q: Can Pydantic AI search for specific Amplitude users?

Yes, the searchusers tool locates accounts by email, user ID, or device ID. Once found, the agent runs getuseractivity to pull that specific individual's product journey.

Q: How does Pydantic AI handle Amplitude event segmentation errors?

The framework validates the inputs for the eventsegmentation tool before sending the request. If the LLM tries to pass a raw timestamp instead of the required YYYYMMDD string, Pydantic throws a runtime error immediately.

Q: What happens if the Amplitude cohort download takes too long?

The getcohort tool is designed to poll the backend. Your Pydantic AI agent will wait for the ready signal instead of crashing or hallucinating a premature response.

Wire Amplitude into Pydantic AI to validate retention metrics, funnel conversions, and revenue analysis at runtime.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Amplitude MCP to Pydantic AI

Create your Vinkius account to connect Amplitude to Pydantic AI and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Amplitude with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Type-Safe Funnels via Pydantic AI

The `get_funnel` tool accepts comma-separated event names to map conversion rates across your product. Pydantic AI forces the LLM to format the YYYYMMDD date parameters perfectly before making the request. If the agent hallucinates a date format, the framework fails loudly with a validation error. Tracking daily income uses the `revenue_analysis` tool. The agent pulls revenue figures for specific time blocks and validates the output against your defined schemas. This model-agnostic approach means you can swap OpenAI for Anthropic without rewriting your financial reporting logic.

Calculate Retention Metrics

Executing the `get_retention` tool tracks how many users perform a start event and later trigger a return event. You get hard numbers on product stickiness rather than guessed approximations. Connecting this MCP Server ensures your agent works with real behavioral data. Grouping these users happens through the `list_cohorts` tool. Your agent pulls the available behavioral segments before deciding which ones need deeper analysis. Downloading the actual members requires the `get_cohort` tool, which handles the necessary polling automatically.

Extract Raw Amplitude Events

Pulling unaggregated data requires the `export_events` tool. The agent specifies a time block using the YYYYMMDDTHH format to grab every tracked action. Because you are using `MCPToolset("http://...")`, the Streamable HTTP transport manages the heavy payload without dropping the connection. Analyzing top-level traffic relies on the `active_users` tool. The agent fetches DAU, WAU, or MAU metrics for your specified date range. Pydantic AI ensures the resulting integer counts map correctly to your internal reporting dashboards.

Setup guide

Set up Amplitude MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "amplitude-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to Amplitude tools.",
)

result = await agent.run("List recent Amplitude transactions")
print(result.output)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Amplitude. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Amplitude now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Amplitude MCP in Pydantic AI

Start with `pip install "pydantic-ai-slim[mcp]"`. Use the unified `MCPToolset("http://...")` class with your Vinkius endpoint. Pass this toolset into your Agent constructor. Do not use the deprecated `MCPServerHTTP` class.

Yes, the `search_users` tool locates accounts by email, user ID, or device ID. Once found, the agent runs `get_user_activity` to pull that specific individual's product journey.

The framework validates the inputs for the `event_segmentation` tool before sending the request. If the LLM tries to pass a raw timestamp instead of the required YYYYMMDD string, Pydantic throws a runtime error immediately.

The `get_cohort` tool is designed to poll the backend. Your Pydantic AI agent will wait for the ready signal instead of crashing or hallucinating a premature response.

Your setup queries raw event timelines containing precise user locations and device specs. Vinkius secures this pipeline using ephemeral containers that spin down the millisecond the request completes. No trace of the query or the response remains on our disks.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript