# Bland AI MCP for AI Agents MCP

> Bland AI gives your AI client full control over automated voice workflows, handling complex inbound and outbound phone calls with natural conversation. You can build persistent agent personas, send calls to multiple countries, and programmatically analyze transcripts for insights.

## Overview
- **Category:** superpower
- **Price:** Free
- **Tags:** ai-voice-agent, automated-calling, conversational-ai, outbound-calling, inbound-support, voice-automation

## Description

Automate your entire call center operation without needing a dedicated telephony team. Bland AI lets you connect your account to any compatible AI client, giving you full control over high-fidelity voice communication. You can use your agent to initiate outbound calls—even orchestrating them to multiple countries—and build persistent AI personas that maintain a consistent brand voice every time they talk. Instead of manually listening through hours of recordings, the system processes finished calls for you, extracting specific variables or sentiment summaries using post-call analysis. It’s all accessible via your existing workflow in the Vinkius catalog. Your agent becomes your dedicated telephony engineer, handling everything from appointment reminders to lead qualification and delivering clean, structured data directly into your process.

## Tools

### analyze_call_transcript
Runs post-call analysis on a transcript to extract specific data points or summaries.

### stop_active_call
Immediately ends an ongoing phone call, regardless of who initiated it.

### create_voice_agent
Builds and saves a persistent AI persona with specific voice settings and instructions.

### delete_voice_agent
Permanently removes an existing, configured AI persona.

### get_agent_config
Retrieves the current settings and instructions for a specific AI agent.

### get_call_details
Fetches complete details, including high-fidelity transcripts, for a given call ID.

### list_voice_agents
Retrieves a list of all configured and managed AI personas.

### list_recent_calls
Lists the metadata for phone calls that were recently completed.

### list_phone_numbers
Provides a list of all purchased and active telephone numbers in your account.

### list_available_voices
Shows the directory of high-fidelity AI voices available for use.

### send_phone_call
Sends an outbound phone call using a specified AI agent and number.

### update_agent_config
Modifies the existing settings, prompts, or instructions for an established AI persona.

## Prompt Examples

**Prompt:** 
```
Send a call to John at 555-123-4567 asking if he can meet next week.
```

**Response:** 
```
**Call Initiated!**

I've sent the outbound call. The agent is briefed: Confirm availability for next week and suggest Tuesday or Thursday afternoon.

*   **Status:** Active (Outbound)
*   **Number:** +1 555-123-4567
*   **Agent Persona:** Sales Associate Alpha

I'll alert you when the transcript is ready for review. Need me to start monitoring for key phrases like 'budget constraints'? [Yes/No]
```

**Prompt:** 
```
Show me all my active agents and their settings.
```

**Response:** 
```
**Agent List (3 Configured)**

Here's what you have set up:

1. **Support Maya:** *Purpose:* Appointment reminders. *Voice:* Female, Calm.
2. **Sales Mason:** *Purpose:* Lead qualification calls. *Voice:* Male, Professional.
3. **Tech Support Bot:** *Purpose:* Password reset flow. *Voice:* Neutral, Direct.

You can view the detailed prompt configuration for any of these agents by asking me to `get_agent_config(agent='Support Maya')`.
```

**Prompt:** 
```
Get the transcript and analysis summary for call ID 987.
```

**Response:** 
```
**Transcript Retrieval:**
*   **Call ID:** 987
*   **Date:** 2026-06-12
*   **Summary:** The client confirmed interest in the premium package but cited budget concerns. **(Analysis Complete)**

**Key Insights Extracted:**
*   **Sentiment:** Neutral/Interested (Score: 7/10)
*   **Actionable Variable:** Needs pricing options for mid-tier packages.
*   **Next Step Suggestion:** Follow up with a quote document via email.
```

## Capabilities

### Initiate targeted calls
Send an outbound phone call using a configured AI agent to specific numbers.

### Build brand personas
Create and manage persistent AI personalities, setting fixed prompts, voices, and core instructions for consistency.

### Manage call records
List recent phone calls or fetch full details and transcripts for any specific interaction.

### Analyze conversation data
Run post-call analysis on transcripts to pull out structured insights, sentiment scores, or key variables.

### Control agent settings
Retrieve or modify the core configuration and instructions for any existing AI persona.

## Use Cases

### Qualifying leads at scale
A sales lead requires verification. Instead of calling manually, the agent uses `send_phone_call` to execute a scripted call asking key qualifying questions. Afterward, using `get_call_details`, your workflow pulls the transcript and sends it for analysis.

### Standardizing support scripts
Your customer service team needs consistency across all calls. You use `create_voice_agent` to build a persona with fixed prompts, ensuring every agent sounds like the brand's voice and knows exactly what information to gather.

### Tracking campaign performance
You need to know if last week’s outreach worked. Your workflow calls `list_recent_calls` to get a list of interactions, then uses `analyze_call_transcript` on the transcripts to generate a summary report showing conversion rates.

### Updating agent behavior mid-campaign
The market shifts and your sales script needs changes. Instead of manually editing documentation, you use `update_agent_config` to modify the core prompts for the AI persona instantly before the next campaign run.

## Benefits

- Stop sifting through hours of audio. After a call, you can run `analyze_call_transcript` to immediately pull out structured data points like sentiment or specific lead variables.
- Maintain perfect brand consistency by creating persistent AI personas using `create_voice_agent`. Your agent will follow fixed prompts and voices every time it calls.
- Manage your entire voice infrastructure programmatically. You can list available numbers (`list_phone_numbers`) and check which high-fidelity voices you own (`list_available_voices`).
- `send_phone_call` lets you trigger outreach programs directly from your workflow, providing real-time instructions to the AI agent during the call.
- If a call goes wrong or needs immediate termination, `stop_active_call` gives you the ability to halt the conversation instantly through your client.

## How It Works

The bottom line is that you connect your preferred AI client once, then treat all voice communication—from setup to analysis—as a single set of commands.

1. Subscribe to this MCP on Vinkius. Then, grab your API Key from your Bland AI dashboard.
2. Start by listing purchased phone numbers or available voices using your AI client.
3. Finally, send an outbound call and use your agent to process the resulting transcripts for actionable data.

## Frequently Asked Questions

**How can Bland AI help me automate my sales calls?**
Bland AI lets your agent make high-fidelity, scripted outbound calls across multiple countries. You simply tell your client the goal, and it handles the conversation while recording everything for later analysis.

**Can I keep my brand voice consistent if I use Bland AI?**
Absolutely. By using the MCP to create a persistent agent persona, you define fixed prompts, voices, and personality settings that every call must adhere to, ensuring total brand consistency.

**Does Bland AI help me find insights from old calls?**
Yes. You can use the MCP's analysis tools on transcripts from past interactions. It pulls out structured data—like sentiment or specific variables—so you don't have to listen to hours of audio.

**Is Bland AI only for sales, or can I use it for support?**
It works for both. You can build a dedicated agent persona specifically for customer service tasks, like automating appointment reminders and collecting satisfaction data after an interaction.

**What if the call needs to stop suddenly?**
The MCP gives you tools that allow you to programmatically halt any ongoing phone call, giving you immediate control over your voice workflows when necessary.