mcp s
A collection of MCP servers to allow LLMs to use the qiskit,qiskit-ibm-runtime library, qiskit-ibm-transpiler, qiskit-code-assistant service library and others
claude mcp add --transport stdio qiskit-mcp-servers uv run qiskit-mcp-server
How to use
This MCP server collection exposes core Qiskit-related MCP servers that enable AI assistants and agents to interact with IBM Quantum services and Qiskit libraries. The core server provides capabilities to generate quantum code, manage circuit construction and transpilation, and supply QASM/QPY serialization utilities, while the surrounding ecosystem includes optional companions for IBM Runtime access, transpilation, code-assisted programming, and community tools. You can start by running the core qiskit-mcp-server via UV, then query the available tools and agents through your AI workspace to generate, optimize, or execute quantum circuits on real or simulated backends. The servers are designed to be invoked from an LLM-enabled workflow, enabling context-aware suggestions, backend analysis, and execution monitoring as part of end-to-end quantum software development and experimentation.
How to install
Prerequisites:
- Python 3.10 or newer
- uv (Astral uv) package manager
- IBM Quantum account and API token (for runtime/backends)
- Optional: access to Qiskit Code Assistant for code-assistant server features
Install from PyPI (recommended):
# Install all MCP servers (core + community)
pip install qiskit-mcp-servers[all]
# Or install just the core servers (default)
pip install qiskit-mcp-servers
Install individual components (examples):
# Qiskit server only
pip install qiskit-mcp-servers[qiskit]
# Code Assistant server only
pip install qiskit-mcp-servers[code-assistant]
# IBM Runtime server only
pip install qiskit-mcp-servers[runtime]
# IBM Transpiler server only
pip install qiskit-mcp-servers[transpiler]
# Qiskit Gym (community) server only
pip install qiskit-mcp-servers[gym]
Running from source (each server runs independently):
# Qiskit Server
cd qiskit-mcp-server
uv run qiskit-mcp-server
# Qiskit Code Assistant Server
cd qiskit-code-assistant-mcp-server
uv run qiskit-code-assistant-mcp-server
# IBM Runtime Server
cd qiskit-ibm-runtime-mcp-server
uv run qiskit-ibm-runtime-mcp-server
Additional notes
Notes and tips:
- Verify Python 3.10+ compatibility; some environments benefit from Python 3.11+.
- Ensure your IBM Quantum API token is configured (e.g., via IBM Qiskit settings or environment variables) if you plan to access real hardware.
- For the Code Assistant and other cloud-backed services, access credentials or licenses may be required; consult the specific server’s documentation for prerequisites.
- If you’re running multiple MCP servers, consider isolating them in separate virtual environments to avoid dependency conflicts.
- Use the MCP registry and examples within each server’s folder to understand common agent patterns and LangChain integrations.
- Logging and debugging can be enhanced by setting verbose environment variables or running with UV’s debugging options when troubleshooting connectivity or backend selection.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP