How to Use the Modelbit (ML Model Deployments) MCP in OpenAI Agents SDK

Q: What happens if my model deployment on Modelbit (ML Model Deployments) returns an error during an OpenAI Agents SDK run?

The getinference tool returns the raw error payload directly to the agent. The agent then reads the error message and can either retry the call with corrected inputs or trigger a handoff.

Q: How does this integration protect my proprietary feature inputs and model outputs?

All feature inputs and model outputs processed by getinference run inside an isolated V8 sandbox on Vinkius. Your raw data never persists on the hosting platform, and connection tokens are kept in ephemeral memory.

Run real-time machine learning predictions directly inside your OpenAI Agents SDK workflows with strict runtime guardrails.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Modelbit (ML Model Deployments) MCP to OpenAI Agents SDK

Create your Vinkius account to connect Modelbit (ML Model Deployments) to OpenAI Agents SDK 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 Modelbit (ML Model Deployments) with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Run predictions inside OpenAI Agents SDK

The `get_inference` tool lets your agent send raw data payloads to your deployed Modelbit endpoints and get back model predictions instantly. You configure this tool once, and your agent handles the serialization and payload formatting without manual API plumbing. This setup eliminates the need to write custom wrapper functions for every new model version. Your agent detects the schema, prepares the feature vector, and triggers the endpoint.

Safe model execution with built-in guardrails

This MCP Server exposes your machine learning models to your agent team with strict execution boundaries. The agent queries `get_inference` only when specific validation rules are met, preventing accidental or malformed model calls. You track every single inference request directly in the OpenAI developer dashboard. Check the logs. If an agent tries to pass incorrect feature types, the system flags the error before hitting your production environment.

Automated feature mapping for agent handoffs

Specialized agents pass prediction tasks to one another using the `get_inference` tool as a shared capability. The routing agent receives raw user text, extracts the necessary parameters, and hands the clean payload to the inference agent. This division of labor prevents model input pollution. By isolating the inference step to a dedicated agent, you maintain clean boundaries between natural language processing and raw numerical model outputs.

Setup guide

Set up Modelbit (ML Model Deployments) MCP in OpenAI Agents SDK

Prerequisites

Python 3.10+ installed
openai-agents package (pip install openai-agents)
Active Vinkius subscription with a valid endpoint token

1

Install the SDK

Run pip install openai-agents to install the OpenAI Agents SDK. The MCP integration is built-in — no extra dependencies needed.
2

Connect via SSE transport

Use MCPServerSse with your Vinkius endpoint URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. The SDK auto-discovers all Modelbit (ML Model Deployments) tools at runtime.
3

Create your Agent

Pass the MCP to Agent(mcp_servers=[server]). The agent receives Modelbit (ML Model Deployments) tools as native definitions — JSON schemas resolve automatically.
4

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Modelbit (ML Model Deployments) tools and returns structured results. Copy the full example on the right to get started.

agent.py

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerSse

async def main():
    async with MCPServerSse(
        url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
    ) as server:
        agent = Agent(
            name="Modelbit (ML Model Deployments) Agent",
            instructions="You have access to Modelbit (ML Model Deployments) tools.",
            mcp_servers=[server],
        )
        result = await Runner.run(agent, "List recent transactions")
        print(result.final_output)

asyncio.run(main())

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 Modelbit. 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 Modelbit (ML Model Deployments) 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 Modelbit (ML Model Deployments) MCP in OpenAI Agents SDK

Install the SDK using pip, then initialize the MCPServerStreamableHttp class with your Vinkius endpoint. This MCP setup auto-discovers the tools and exposes them when constructing your agent.

Yes. Set cacheToolsList=True in your server configuration to avoid fetching the tool schema on every single agent interaction. This reduces latency when calling `get_inference`.

The `get_inference` tool returns the raw error payload directly to the agent. The agent then reads the error message and can either retry the call with corrected inputs or trigger a handoff.

No. The MCP Server handles the schema translation automatically. Your agent reads the tool definition and formats the input arguments to match what your Modelbit deployment expects.

All feature inputs and model outputs processed by `get_inference` run inside an isolated V8 sandbox on Vinkius. Your raw data never persists on the hosting platform, and connection tokens are kept in ephemeral memory.

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