How to Use the LangSmith MCP in Pydantic AI

Q: Does Pydantic AI validate the response of LangSmith tools?

Yes, every telemetry payload from langsmithgetrun is validated at runtime. If the API returns unexpected trace fields, the framework raises a validation error instead of failing silently.

Bring type-safe LangSmith telemetry into your Pydantic AI agents via this MCP Server to catch errors.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect LangSmith MCP to Pydantic AI

Create your Vinkius account to connect LangSmith 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 LangSmith with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Type-Safe Tracing with Pydantic AI

Pydantic AI forces your agent to validate all incoming data structures at runtime. This MCP Server exposes telemetry tools that return strictly structured data, so your agent never gets confused by messy logs. When your agent calls `langsmith_list_runs`, the response is parsed against strict schemas. If a trace format changes, the system catches it immediately rather than letting corrupted data slip into your monitoring pipelines.

Audit Projects with Type Validation

Monitoring multiple staging environments requires clean aggregate metrics. You cannot rely on loose JSON structures when auditing performance across your entire system. Use `langsmith_list_projects` to fetch validated run counts, feedback metrics, and median latencies. This ensures your auditing scripts receive clean, typed data every single time.

Inspect Specific Run Details

When a specific agent action fails, you need to pull the exact inputs and outputs of that run. This requires a direct query to the trace database. By calling `langsmith_get_run`, you can retrieve the full details of any execution. The returned payload is structured, making it easy to parse and display within your terminal or debugging interface.

Setup guide

Set up LangSmith 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": {
        "langsmith-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

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

result = await agent.run("List recent LangSmith 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 LangSmith. 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 LangSmith 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 LangSmith MCP in Pydantic AI

Use the unified MCPToolset class with your server's HTTP endpoint. Pass this toolset to your agent's constructor to let it execute `langsmith_list_runs` with built-in schema validation.

Yes, every telemetry payload from `langsmith_get_run` is validated at runtime. If the API returns unexpected trace fields, the framework raises a validation error instead of failing silently.

Yes, this setup requires an externally running server. You connect to this MCP setup using the SSE or Streamable HTTP transport, allowing your agent to query `langsmith_list_projects` seamlessly.

Yes. Because Pydantic AI is model-agnostic, you can trace executions from local models or commercial APIs and pull those metrics through the same interface.

This server processes your LangSmith trace metadata, token counts, and latencies within a secure, zero-trust V8 sandbox. No analytical data is stored on our servers, and all requests are executed using ephemeral tokens.

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