How to Use the Deepgram MCP in OpenAI Agents SDK

Q: How do I handle Deepgram transcription errors in OpenAI Agents SDK?

When transcribeaudiourl fails due to a bad link, the SDK catches the exception and routes it through your agent's error-handling loop. You can configure a fallback agent to ask the user for a new URL or try a different audio format.

Q: Can OpenAI Agents SDK manage multiple Deepgram projects dynamically?

Yes, the agent can call listdeepgramprojects to discover active workspaces and switch keys accordingly. The SDK's tracing dashboard will log each project switch so you can audit which agent accessed which workspace.

Q: How do I cache Deepgram tool definitions in OpenAI Agents SDK?

Set the cacheToolsList parameter to true when configuring your server streamable HTTP connection. This prevents the SDK from querying the MCP Server for tool schemas on every single agent turn, saving network roundtrips.

Q: How does the agent convert text to speech asynchronously?

The python SDK runs the converttexttospeech tool inside an async context manager. This keeps your main event loop free to process other tasks while the audio file is being generated.

Build production-ready OpenAI Agents SDK setups that run this MCP Server to transcribe audio URLs and run TTS 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 Deepgram MCP to OpenAI Agents SDK

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

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Safe Audio Processing for OpenAI Agents SDK

Your autonomous agents can now convert text to voice or read audio files without unchecked execution risks. By combining this MCP Server with OpenAI's dashboard tracing, you watch every call to `transcribe_audio_url` happen in real-time, giving you total visibility over raw audio inputs before they hit your agent's context. The SDK handles the heavy lifting of agent handoffs. When a customer support agent hears a user's voice file, it spins up a specialized transcription agent, runs the tool, and routes the text back to the main loop with zero manual wiring.

Guarded API Resource Tracking

Stop worrying about runaway API costs when deploying autonomous loops. This MCP integration lets your agent check its own burn rate using `get_project_usage` to pause loops or switch models dynamically if budgets are exceeded. Before your Python agent executes `list_api_keys`, the built-in SDK guardrails validate the action against your security policy. If a rogue prompt tries to fish for credentials, the SDK blocks the call before it ever contacts the Deepgram endpoint.

Automated Model Selection and TTS

Let your agents select the best voice model on the fly. By calling `list_available_models`, the agent checks what high-performance options are live and immediately feeds that choice into `convert_text_to_speech` for instant voice generation. This setup runs entirely within an async context manager. You get non-blocking audio generation that keeps your production agent systems responsive, even when handling massive text payloads.

Setup guide

Set up Deepgram 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 Deepgram tools at runtime.
3

Create your Agent

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

Run the agent

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

When `transcribe_audio_url` fails due to a bad link, the SDK catches the exception and routes it through your agent's error-handling loop. You can configure a fallback agent to ask the user for a new URL or try a different audio format.

Yes, the agent can call `list_deepgram_projects` to discover active workspaces and switch keys accordingly. The SDK's tracing dashboard will log each project switch so you can audit which agent accessed which workspace.

Set the `cacheToolsList` parameter to true when configuring your server streamable HTTP connection. This prevents the SDK from querying the MCP Server for tool schemas on every single agent turn, saving network roundtrips.

The python SDK runs the `convert_text_to_speech` tool inside an async context manager. This keeps your main event loop free to process other tasks while the audio file is being generated.

Your raw audio URLs and generated transcripts pass directly through this MCP Server inside ephemeral sandboxes. No persistent storage is used, meaning your transcripts exist only long enough for the OpenAI agent to process them.

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