Build and deploy
MCP Agents

Open-source SDKs and infratructure for MCP that developers love.

Or use our open source libraries

pip install mcp-use
npx create-mcp-use-app

Library used by devs from

6sense Logo
Elastic Logo
IBM Logo
Innovacer Logo
Intuit Logo
NVIDIA Logo
Oracle Logo
Red Hat Logo
Tavily Logo
Verizon Logo
6sense Logo
Elastic Logo
IBM Logo
Innovacer Logo
Intuit Logo
NVIDIA Logo
Oracle Logo
Red Hat Logo
Tavily Logo
Verizon Logo
Features

The stack for MCP development

Build, preview, ship, iterate with mcp-use.

Without SDK98 lines
1class MCPClient:
2 def __init__(self):
3 # Initialize session and client objects
4 self.session: Optional[ClientSession] = None
5 self.exit_stack = AsyncExitStack()
6 self.anthropic = Anthropic()
7 # methods will go here
8
9async def connect_to_server(self, server_script_path: str):
10 is_python = server_script_path.endswith('.py')
11 is_js = server_script_path.endswith('.js')
12 if not (is_python or is_js):
13 raise ValueError("Server script must be a .py or .js file")
14
15 command = "python" if is_python else "node"
16 server_params = StdioServerParameters(
17 command=command,
18 args=[server_script_path],
19 env=None
20 )
21
22 stdio_transport = await self.exit_stack.enter_async_context(stdio_client(server_params))
23 self.stdio, self.write = stdio_transport
24 self.session = await self.exit_stack.enter_async_context(ClientSession(self.stdio, self.write))
25
26 await self.session.initialize()
27
28 # List available tools
29 response = await self.session.list_tools()
30 tools = response.tools
31 print("\nConnected to server with tools:", [tool.name for tool in tools])
32 async def process_query(self, query: str) -> str:
33 """Process a query using Claude and available tools"""
34 messages = [
35 {
36 "role": "user",
37 "content": query
38 }
39 ]
40
41 response = await self.session.list_tools()
42 available_tools = [{
43 "name": tool.name,
44 "description": tool.description,
45 "input_schema": tool.inputSchema
46 } for tool in response.tools]
47
48 # Initial Claude API call
49 response = self.anthropic.messages.create(
50 model="claude-3-5-sonnet-20241022",
51 max_tokens=1000,
52 messages=messages,
53 tools=available_tools
54 )
55
56 # Process response and handle tool calls
57 final_text = []
58
59 assistant_message_content = []
60 for content in response.content:
61 if content.type == 'text':
62 final_text.append(content.text)
63 assistant_message_content.append(content)
64 elif content.type == 'tool_use':
65 tool_name = content.name
66 tool_args = content.input
67
68 # Execute tool call
69 result = await self.session.call_tool(tool_name, tool_args)
70 final_text.append(f"[Calling tool {tool_name} with args {tool_args}]")
71
72 assistant_message_content.append(content)
73 messages.append({
74 "role": "assistant",
75 "content": assistant_message_content
76 })
77 messages.append({
78 "role": "user",
79 "content": [
80 {
81 "type": "tool_result",
82 "tool_use_id": content.id,
83 "content": result.content
84 }
85 ]
86 })
87
88 # Get next response from Claude
89 response = self.anthropic.messages.create(
90 model="claude-3-5-sonnet-20241022",
91 max_tokens=1000,
92 messages=messages,
93 tools=available_tools
94 )
95
96 final_text.append(response.content[0].text)
97
98 return "\n".join(final_text)
With mcp-use SDK8 lines
1# Connect to remote server pool
2client = MCPClient(server_pool="marketing_servers")
3# Use any model you like
4llm = AnyModelProvider(model="any-4o", api_key=...)
5# Create agent with tools
6agent = MCPAgent(llm=llm, client=client, max_steps=30)
7# Run the query
8result = await agent.run("Create a post about our latest release")
90 lines saved
Build

Full-Stack MCP Framework

  • MCP Agent 6 Lines of Code: That's all it takes to spin up a tool-using agent in Python or TypeScript.
  • MCP Apps in ChatGPT: Render interactive UI components directly in chat clients like ChatGPT.
  • MCP Server: Scaffold a production-ready server instantly and take inspo from templates.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
import { createMCPServer } from 'mcp-use/server'
// The MCP Inspector is automatically mounted at /inspector
const server = createMCPServer('my-mcp-server', {
version: '1.0.0',
description: 'An MCP server with Apps SDK support for ChatGPT',
baseUrl: process.env.MCP_URL,
})
// UI Widgets are React components in the "resources/" folder.
// They are automatically registered as both MCP tools and resources.
// ...
// Add your tools, resources, and prompts here
MCP Server
mcp-useInspector
Select a MCP Tool to inspect
Preview

Iterate with inspector

  • Built-in Debugger: Every server launches with a visual Inspector to test tools, edit prompts, and view resources in your browser.
  • Real-time Traffic: View raw JSON-RPC messages to see exactly what the LLM sent and what your tool returned.
  • Sandbox Testing: Validate tool execution safely before connecting to a live agent.

Stdio MCP servers

Hosted short-lived MCP servers

VM

Third-party MCP servers

External remote services

Cloud

Your MCP servers

Managed or self-hosted

mcp-use hosting

Ship

Server deploy and agent runtime

  • Deploy MCP Servers with GitHub App: Install our GitHub App to automatically build and deploy your server on every push.
  • Isolated Scaling: Scale horizontally to handle thousands of concurrent users and prevet race conditions.
  • Agent Persistent Context: Agents usually forget everything. Ours do not. The runtime maintains state and memory across sessions without needing an external database.
Authentication
Profiles & ACL
MCP Configs
Envs & Pools
mcp-use gateway
MCP servers routing
Cache
Skills & Memory
Monitor
Observability
Iterate

Control Plane for MCP

  • Observability & Learning: See how users interact with your MCP servers. Analyze requests and let agents retain skills over time.
  • Auth & Access Control: Security is built-in. Use role-based controls and profiles to manage permissions across employees/clients/environments.
  • Centralized Config: Manage all configurations from one dashboard and import them directly via the SDK.
Offering

The control plane for MCP

Need to run on your stack? Deploy to your existing infrastructure.
Prefer a managed solution? Use mcp-use Cloud for zero setup and overheads.

MCP Apps & AI Agents
ChatGPT
Your AI product
MCP client
pip install mcp-use
npx create-mcp-use-app
Claude (Desktop / Code)
Authentication
Profiles & ACL
MCP Configs
Envs & Pools
mcp-use gateway
MCP servers routing
Cache
Skills & Memory
Monitor
Observability
MCP Servers
Third-party MCP servers
External remote services
Cloud
Your MCP servers
Managed or self-hosted
mcp-use hosting
Stdio MCP servers
Hosted short-lived MCP servers
VM
What you can build

Anything you do with a web app
you can do with mcp-use

Start building with mcp-use and see how we can help you.

MCP Apps (for ChatGPT)

MCP Apps (for ChatGPT)

Reach millions of users on ChatGPT and Claude. Display your UI interfaces directly in chat clients with a simple React component.

MCP Servers

MCP Servers

Expose your services to agents using the open MCP standard. This allows any AI agent to understand and use your services reliably.

MCP Agents

MCP Agents

Build MCP agents in minutes. You can expose them immediately via a pre-built chat interface or integrate them into your own codebase with just 6 lines of code.

Testimonials <3

Developers love

Thousands of dev-teams are building with mcp-use

Been thinking about trying your @mcpuse today when I started building a ChatGPT app, but wasn't sure if it's the right use case. Now I assume it is 🙂

Paolo Perazzo
Paolo Perazzo@SiVola

Great addition to the generic MCP Client collection.

Kingsley Uyi Idehen
Kingsley Uyi Idehen@kidehen

Infra isn't just plumbing. It's distribution. Examples: AgentMail: inboxes for agents, mcp-use: MCP infra, DeepAware: RL for data center ops. Whoever owns the rails agents use to talk to the outside world ends up with Twilio-style power.

Vedika Jain
Vedika Jain@vedikaja_in

mcp-use is by far the best Python framework for building an agent MCP. It's so easy to set up an agent connected to MCP servers through the command line, that way I can quickly iterate on my MCP servers and test. Most people are using MCP servers in chat clients like Cursor / Claude Code, but MCP server use by agents will explode in the near future. I think it's important to test your server in an agent environment, and mcp-use provides that.

Matt Wang
Matt Wang@mcpjams

MCP-Use is the open source way to connect any LLM to any MCP server and build custom agents that have tool access, without using closed source or application clients.

Adam Silverman
Adam Silverman@adamsilverman

that's why we love open-source! @NASA is building an agent with MCP using our library @mcpuse 🚀

pederzh
pederzh@pederzh

Been thinking about trying your @mcpuse today when I started building a ChatGPT app, but wasn't sure if it's the right use case. Now I assume it is 🙂

Paolo Perazzo
Paolo Perazzo@SiVola

Great addition to the generic MCP Client collection.

Kingsley Uyi Idehen
Kingsley Uyi Idehen@kidehen

Infra isn't just plumbing. It's distribution. Examples: AgentMail: inboxes for agents, mcp-use: MCP infra, DeepAware: RL for data center ops. Whoever owns the rails agents use to talk to the outside world ends up with Twilio-style power.

Vedika Jain
Vedika Jain@vedikaja_in

mcp-use is by far the best Python framework for building an agent MCP. It's so easy to set up an agent connected to MCP servers through the command line, that way I can quickly iterate on my MCP servers and test. Most people are using MCP servers in chat clients like Cursor / Claude Code, but MCP server use by agents will explode in the near future. I think it's important to test your server in an agent environment, and mcp-use provides that.

Matt Wang
Matt Wang@mcpjams

MCP-Use is the open source way to connect any LLM to any MCP server and build custom agents that have tool access, without using closed source or application clients.

Adam Silverman
Adam Silverman@adamsilverman

that's why we love open-source! @NASA is building an agent with MCP using our library @mcpuse 🚀

pederzh
pederzh@pederzh

Join MCP community

Get help, share your projects, and get inspired.

The community for developers building with MCP and mcp-use.

Stats

Among the most popular MCP projects

mcp-use is one of the most popular open-source MCP projects

1.5M+

MCP agents executed with mcp-use

4,000+

Companies used mcp-use SDKs to experiment and build MCP-enabled agents and MCP servers.

8,000+

Stars on GitHub.

Begin your MCP journey
in the
fastest way

Build with mcp-use SDKs, preview with our inspector, deploy with mcp-use Cloud.