How to Use the Cohere MCP in OpenAI Agents SDK

Q: Can I restrict which Cohere models my OpenAI Agents SDK client can access?

Yes. Use the listmodels tool inside your OpenAI Agents SDK system to check available models like command-r or rerank-v3.5. You can then write a guardrail function that blocks any agent from calling chat or rerank with unauthorized model IDs.

Build production-grade OpenAI Agents SDK systems that run Cohere models via MCP Server with built-in guardrails.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Cohere MCP to OpenAI Agents SDK

Create your Vinkius account to connect Cohere 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 Cohere with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Multi-Agent Cohere Routing with OpenAI Agents SDK

The `chat` tool lets your OpenAI Agents SDK coordinate multi-agent handoffs while routing prompts to Command models. When a user asks for a summary, the supervisor agent calls `chat` using `command-r-plus` to generate the response, passing the output straight to the next specialized agent. Because the SDK tracks every step, you can trace the entire multi-model conversation path directly on your OpenAI dashboard. We configure this by passing the streamable HTTP MCP server instance into the `mcp_servers` list during agent initialization. Set `cacheToolsList=True` to prevent unnecessary roundtrips, keeping your production agent pipelines fast and predictable.

Search Document Reordering via MCP Server

The `rerank` tool optimizes how your OpenAI Agents SDK handles large-scale retrieval-augmented generation. Instead of stuffing raw search results into the agent context, the agent invokes `rerank-v3.5` to calculate exact relevance scores for each document. The agent then drops low-scoring items before they hit the LLM context window. This approach cuts down on token waste and improves response accuracy during complex multi-agent workflows. The SDK validates these tool calls against your guardrails, ensuring the retrieved documents do not bypass safety rules.

Token Budget Enforcement

The `tokenize` tool gives your OpenAI Agents SDK exact token counts before executing expensive API calls. Your supervisor agent uses this tool to measure input texts, then uses `detokenize` to debug any malformed strings during complex agent handoffs. By validating these token lengths before calling the main model, your agent prevents context overflows in production. It keeps your operational costs predictable while running multiple parallel Cohere runs.

Setup guide

Set up Cohere 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 Cohere tools at runtime.
3

Create your Agent

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

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Cohere 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="Cohere Agent",
            instructions="You have access to Cohere 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 Cohere. 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 Cohere 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 Cohere MCP in OpenAI Agents SDK

Run pip install openai-agents and instantiate MCPServerStreamableHttp pointing to the Vinkius endpoint. Pass this server object into your Agent constructor using the mcp_servers parameter to auto-discover all six Cohere tools.

Yes. Use the list_models tool inside your OpenAI Agents SDK system to check available models like command-r or rerank-v3.5. You can then write a guardrail function that blocks any agent from calling chat or rerank with unauthorized model IDs.

Yes. The SDK natively supports async context managers. When your agent calls the embed tool with embed-v4 to generate vectors for a batch of texts, the HTTP request runs asynchronously, keeping your multi-agent execution pipeline unblocked.

Your agent uses the chat tool's tool-calling parameters to request structured data. The agent receives the tool calls in the response, executes them locally, and sends the results back to the Cohere model to complete the loop.

Vinkius runs the server in an ephemeral, zero-trust V8 Isolate sandbox. Your raw text payloads, embeddings, and token IDs are processed in isolation and sent directly to the Cohere API, with no data stored locally on Vinkius servers.

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