Bland AI MCP. Run automated calls and manage call data flow.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Bland AI. Automate phone calls and manage agent interactions through your AI client. Send outbound calls, list available agents, and pull full transcripts directly from the conversation.
Manage call workflows—from initial dialing to final analysis—all without touching a phone or dialer. It's the automation layer for sales and support operations.
What your AI agents can do
Get agent
Retrieves specific details about an AI agent.
Get call
Gets detailed metadata for a specific call event.
Get pathway
Retrieves the details of a defined conversation flow.
Send an AI-driven outbound call using send_call and stop it mid-stream with stop_call.
List all available AI agents using list_agents and get details on a specific agent using get_agent.
List recent calls with list_calls and retrieve full transcripts and metadata using get_call and get_transcript.
View available conversation paths with list_pathways and retrieve details on a specific path using get_pathway.
View all available voices with list_voices to select the perfect voice for your brand.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Bland AI MCP Server: 10 Tools for Call & Agent Control
Manage call lifecycles, retrieve full transcripts, and control agent operations using these ten tools.
019d755eget agent
Retrieves specific details about an AI agent.
019d755eget call
Gets detailed metadata for a specific call event.
019d755eget pathway
Retrieves the details of a defined conversation flow.
019d755eget transcript
Gets the full conversation transcript for a finished call.
019d755elist agents
Lists all available AI agents in your account.
019d755elist calls
Lists recent AI calls, providing basic call metadata.
019d755elist pathways
Lists all predefined conversation pathways available.
019d755elist voices
Lists every AI voice available for use in calls.
019d755esend call
Initiates and sends an AI-powered phone call to a given number.
019d755estop call
Stops an active or scheduled phone call immediately.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Bland AI, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
You're gonna run your whole sales or support operation through your AI client. This server lets you send AI-powered calls and manage the whole agent interaction process. You can initiate a call using send_call and stop it dead in its tracks with stop_call. You'll find all your available AI agents by running list_agents, and you can drill down into any specific agent's details using get_agent.
To see what kind of calls you've been running, you can check out recent activity with list_calls, and you can grab detailed info on any specific call using get_call. When a call wraps up, you've got the full transcript waiting for you via get_transcript. You can also build out complex conversation logic by listing all available conversation paths with list_pathways, and you can pull up the specifics of a flow with get_pathway.
Need a different tone? You can see every voice available for calls using list_voices. You'll manage all this—from the initial dial to the final analysis—without ever touching a phone or a dialer.
How Bland AI MCP Works
- 1 Subscribe to the server and plug in your Bland AI API Key.
- 2 Connect your AI client to the MCP server.
- 3 Your agent calls the necessary tools (e.g.,
list_agentsthensend_call) to manage the entire phone interaction.
The bottom line is, you talk to your AI client, and it uses the Bland AI tools to run the phone calls and pull the data.
Who Is Bland AI MCP For?
The operations lead who gets sick of manually dialing follow-up calls. The sales manager who needs to know exactly what happened in a call without logging into a separate CRM. Or the developer building a custom agent that needs reliable, repeatable phone interaction capabilities.
Runs automated follow-up campaigns and appointment reminders, eliminating the need for manual dialing.
Monitors call performance and accesses full call transcripts to train or adjust support agents.
Integrates phone call logic into custom agent workflows, handling state and data retrieval programmatically.
What Changes When You Connect
- Send automated calls with
send_calland customize the experience by selecting the right voice vialist_voices. You handle the campaign; the AI handles the dialing. - Instantly monitor performance. Use
list_callsto see recent activity, then callget_callto pull deep metadata and call summaries. - Deep dive into conversations. Call
get_transcriptto get the full text log of any completed call, allowing you to analyze outcomes later. - Control the conversation. Use
get_pathwayandlist_pathwaysto enforce specific, complex logic (like 'if X, ask Y') into the call flow. - Manage your assets. Use
list_agentsandget_agentto track which AI 'personas' are available and how they perform. - Operational safety net. If a call goes off script or needs stopping,
stop_calllets you cut the connection immediately.
Real-World Use Cases
Running a large-scale follow-up campaign
The marketing team needs to remind 500 leads about a webinar. Instead of manually dialing, the agent calls send_call for each lead. After the campaign, the agent calls list_calls and then get_transcript on the successful calls to build a report.
Debugging a complex sales conversation
A sales call went sideways. Instead of asking a human to listen to the recording, the agent calls list_calls to find the call ID, then get_transcript to get the text log. The agent can then analyze the conversation flow and identify where the logic failed.
Onboarding a new support agent persona
A company hires a new support bot. The agent calls list_agents to see what exists, then uses get_agent to check the configuration and test the bot using send_call with a specific voice from list_voices.
Enforcing multi-step qualification logic
The process requires the AI to ask 3 specific questions in order. The developer configures a pathway using get_pathway. When the call runs, the agent uses this pathway ID to ensure the conversation follows the exact required logic.
The Tradeoffs
Trying to manually copy transcripts
A user watches a recording of a call and tries to copy and paste the dialogue into a spreadsheet. This is time-consuming and often misses critical non-verbal cues.
→
Let your agent call get_transcript directly. It pulls the full, structured conversation log, which you can then pass to a data validation tool for immediate processing.
Forgetting to check agent status
Starting a call with an outdated or incorrect AI agent profile, resulting in the conversation going off track or sounding generic.
→
Always call list_agents first. This ensures you use the latest, properly configured agent for the job.
Handling multiple calls sequentially
Writing a massive script that tries to call get_call for 50 different call IDs and then tries to loop through the results manually.
→
Use list_calls first to get a list of IDs. Then, loop through those IDs, calling get_call for each one in a structured loop.
When It Fits, When It Doesn't
Use this server if your primary goal is automating the entire lifecycle of phone communication: from sending the initial call to retrieving the final, structured data. You need to manage call state, track agent performance, and pull full transcripts. Don't use it if you just need a simple API to make a single call without tracking or managing the state. If you only need to read a single piece of data, you might only need get_transcript. But if you need to do something—like send a call or check a status—you need the whole suite. It's built for workflows, not single actions.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Bland AI. 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.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Dialing out and tracking results shouldn't feel like a full-time job.
Today, running a simple follow-up campaign means juggling a dialer, a CRM, and a spreadsheet. You have to manually log the call outcome, find the recording, and copy the key details into the right column. It's a painful, multi-step process that burns hours just on administration.
With the Bland AI MCP Server, you let your agent handle the dialing. You just call `send_call` and set the parameters. When the call ends, your agent calls `get_transcript` and gives you the clean, structured log. You get the data, instantly.
Bland AI MCP Server: Control the entire call flow.
Manual call monitoring means logging into a dashboard, finding the call ID, and hoping the metadata is still there. You have to check if the call was completed, what the outcome was, and where the conversation went.
Now, your agent uses `list_calls` to see recent activity, then calls `get_call` to get the full metadata, and `get_pathway` to see exactly which logic path the AI followed. The state is visible, and the whole process is controlled.
Common Questions About Bland AI MCP
How do I use the `send_call` tool with Bland AI? +
You call send_call and provide the target phone number and the specific task you want the AI to perform. The agent then initiates the call and returns a Call ID for tracking.
What is the difference between `list_calls` and `get_call`? +
list_calls gives you a quick list of recent calls and basic IDs. get_call pulls the full, detailed metadata for one specific call ID.
Can I get a transcript using the `get_transcript` tool? +
Yes. If a call is finished, calling get_transcript retrieves the full, structured conversation log, showing who said what.
How do I manage agent behavior using `get_agent`? +
You use get_agent to retrieve the current configuration and details of a specific AI persona, allowing you to confirm it's set up correctly before calling it.
Do I need to use `list_voices` before calling `send_call`? +
While not strictly required for the API call, it’s best practice to call list_voices first. This lets you confirm the exact voice ID you need to match your brand's tone.
How do I use `list_agents` to see all the AI personas available? +
The list_agents tool returns a list of all active AI agents. Each agent entry includes its unique ID and a brief description of its function. You can then pass the specific agent ID to get_agent for detailed settings.
If a call fails, how can I use `get_call` to check the error details? +
The get_call tool provides comprehensive metadata for any call, including status codes and error messages if the call did not complete successfully. This lets you quickly diagnose why a call failed without needing manual logs.
What is the purpose of `stop_call` and when should I use it? +
stop_call immediately ends any active or scheduled phone call. Use this tool if you need to intervene mid-call, cancel a scheduled outreach, or correct a mistake.
Can I send a phone call with specific instructions using the agent? +
Yes! Use the send_call action with the target phone number and a 'task' string describing exactly what the AI should say and do during the call.
How do I retrieve the transcript of a completed call? +
Simply ask the agent to get_transcript and provide the Call ID. It will retrieve the full conversation log from the Bland AI engine.
Can I list all available voices before starting a call? +
Yes. Use the list_voices tool to see all AI voices available in your Bland AI account, including their names and unique IDs.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
AutoGen
Orchestrate Microsoft AutoGen multi-agent workflows — manage sessions, agent roles, workflows, and monitor execution logs from any AI agent.
Midjourney AI (Generative Image Arts)
Generate professional AI art via Midjourney — use 'imagine' for text-to-image, upscale grids, and perform camera edits.
Modal (Serverless AI Infrastructure)
Manage serverless compute via Modal — audit active apps, track GPU deployments, and monitor network volumes.
You might also like
Clientify
Manage sales and marketing via Clientify CRM — track contacts, monitor deals, and audit CRM activities directly from any AI agent.
Learn Amp
Combine learning, engagement, and performance in one people development platform that helps employees grow and organizations thrive.
Nifty
Combine tasks, timelines, docs, and chat in one project management workspace that replaces multiple tools at once.