# Deepgram MCP for AI Agents AI Agent Connect

> Deepgram MCP lets you manage and run speech-to-text and text-to-speech workflows directly from your AI agent. You can transcribe audio URLs, generate high-fidelity voices, track your usage limits, and manage your API keys or team members without switching tabs. It connects your Deepgram account to your favorite AI tools so you can build and monitor audio products faster.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_TncgUjBvIqBaRxaOCzAMTEFPrPacCVhmUvpvUOyU/ai-agent-connect
- **Tags:** speech-to-text, text-to-speech, transcription, diarization, audio-processing, neural-networks

## Description

Instead of jumping between browser tabs to check transcription logs or manually creating new API keys, you can handle your entire audio pipeline through a conversation. This Connector gives your agent the hands it needs to actually do the work. You can feed it audio URLs to get instant transcriptions or turn blocks of text into natural-sounding speech. Beyond the core audio tasks, it handles the heavy lifting of account management. You can see exactly how much of your budget is being spent on bytes or minutes, create new access boundaries for different projects, and manage who on your team has permission to see what. It's about moving from 'I need to check the dashboard' to 'Just tell me the transcription status.' By adding this to your setup via the Vinkius catalog, you turn your AI into a functional operator for your Deepgram account rather than just a chatbot that talks about audio.

## Tools

### list_api_keys
List active API keys

### get_project_usage
Check API usage and limits

### list_available_models
List high-performance AI models

### list_deepgram_projects
List your Deepgram projects

### convert_text_to_speech
Generate audio from text (TTS)

### transcribe_audio_url
Transcribe an audio file via URL

## Prompt Examples

**Prompt:** 
```
Transcribe this audio: https://example.com/recording.mp3 using nova-2
```

**Response:** 
```
Transcription complete! Using the Nova-2 model, I've processed the audio stream. Here is the text: **'Welcome to the deep dive into AI orchestration...'**

I also checked the confidence scores for you:
* **Word 1**: 0.99
* **Word 2**: 0.98
* **Word 3**: 0.97
```

**Prompt:** 
```
Generate speech for: 'The future of AI is agentic' using aura-asteria-en
```

**Response:** 
```
Audio stream generated! I've synthesized the text using the Aura Asteria voice. 

**Details:**
* **Status**: Success
* **Voice**: Asteria
* **Format**: High-fidelity binary stream

You can play or download the resulting data now.
```

**Prompt:** 
```
Show me my Deepgram usage for this month
```

**Response:** 
```
Retrieving usage data for your project... 

| Metric | Usage |
| :--- | :--- |
| **Transcription** | 450 minutes |
| **TTS Content** | 12MB |

Your project limits are well within the current thresholds.
```

## Capabilities

### Transcribe audio from URLs
Convert audio files into text using the Nova-2 model.

### Generate high-fidelity speech
Turn raw text into natural sounding audio using Aura voices.

### Monitor credit balances
Check your remaining credits for any project instantly.

### Create scoped API keys
Generate new access keys with specific permissions and comments.

### Manage team permissions
Invite new members or view current project access in one command.

### Track usage statistics
Pull transcription and TTS usage data for specific date ranges.

## Use Cases

### Audio QA Audits
A QA engineer asks the agent to transcribe 10 URLs and check the Nova-2 accuracy scores in a table.

### Budget Tracking
A PM asks for a summary of transcription minutes spent in the last 30 days to plan next month's budget.

### Team Onboarding
A lead dev tells the agent to invite three new engineers to the production speech project.

### Rapid Prototyping
A developer generates a voiceover for a UI demo using a specific Aura voice directly from a text prompt.

## Benefits

- Use `transcribe_url` to turn audio files into text instantly without manual downloads.
- Generate audio via `speak_text` to create lifelike speech for your apps on the fly.
- Use `get_balances` and `get_usage` to stop overspending before it happens.
- Manage security by using `create_key` and `delete_key` to keep your project boundaries safe.
- Use `list_members` and `send_invite` to manage who can access your audio projects.

## How It Works

The bottom line is you get a direct line to Deepgram's audio engine and account controls inside your chat.

1. Get your API key from the Deepgram Console under Settings.
2. Plug the key into your AI client via Vinkius.
3. Ask your agent to transcribe a file or check your project balance.

## Frequently Asked Questions

**Can Deepgram MCP transcribe my audio files?**
Yes, it allows your agent to take audio URLs and turn them into text using the high-speed Nova-2 model.

**How do I manage my Deepgram API keys with this?**
You can ask your agent to list, create, or delete keys for specific projects directly in your chat.

**Can I see my account balance using the Deepgram MCP?**
Yes, the Connector can check your current credit balance and usage stats for any project you have connected.

**Does this support different voices for speech generation?**
Yes, you can use the Aura voices to generate speech from text via your agent for high-fidelity results.

**Can I add team members to my project via the Connector?**
You can use the Connector to list current members and send invites to new people without leaving your chat interface.

**Is this for transcription or just speech?**
It does both. It handles speech-to-text for audio files and text-to-speech for generating voices from text.

**Can my agent transcribe an audio file from a public URL?**
Yes. Use the 'transcribe_url' tool. Provide the public URL of the audio file (WAV, MP3, etc.) and specify the model (e.g., 'nova-2'). The agent will dispatch the request to Deepgram and return the transcribed text instantly.

**How do I generate speech from text using the agent?**
Use the 'speak_text' tool. Provide the text script and the target voice model (e.g., 'aura-asteria-en'). Your agent will trigger the high-fidelity Aura voice engine and return the binary audio stream data.

**Can I monitor my remaining project balance via chat?**
Absolutely. Use the 'get_balances' tool with your project ID. The agent will retrieve your current wallet thresholds and funding limits directly from Deepgram to ensure your audio pipelines stay active.

**How do I get a Deepgram API Key?**
Log in to the Deepgram Console, navigate to the **API Keys** section, and create a new key with the necessary permissions.

**What is the Nova-3 model?**
Nova-3 is Deepgram's latest state-of-the-art transcription model, offering unmatched speed and accuracy for real-world audio.

**Can I synthesize speech in different voices?**
Yes! The `convert_text_to_speech` tool allows you to specify models like `aura-asteria-en` or `aura-orion-en` for different vocal profiles.