</>MCP Agents Market
MCP Server

aiq

by NVIDIA-AI-Blueprints863PythonUpdated 2026-09-02

The AI-Q NVIDIA Blueprint is an open reference example for building intelligent AI agents that connect to your enterprise data, reason using state-of-the-art models, and deliver trusted business insights.

Claude DesktopCursorAny MCP-compatible client

NVIDIA AI-Q Blueprint is an MCP server and enterprise research platform that delivers both quick, cited answers and comprehensive research reports using NVIDIA's state-of-the-art models. Built on NVIDIA NeMo Agent Toolkit and LangChain Deep Agents, it orchestrates multi-phase research workflows connecting to enterprise data sources including web search, academic papers, and knowledge layers. The system supports shallow research for fast queries and deep research with concurrent workers, structured planning, and citation-backed outputs. Developers can deploy AI-Q as a standalone MCP server exposing submit/poll/report operations, integrate it via REST API with async job support, or run it through CLI and web interfaces.

Key Features

Standalone MCP server with stateless submit_query, poll_query, and get_final_report operations over Streamable HTTP
Orchestrated research workflows with intent classification routing between shallow (fast, cited) and deep (report-style) research modes
Structured deep research with advisory source routing, concurrent researcher workers, bounded tool batching, and dedicated writer agents
Multi-source integration including Tavily, Exa, You.com, Nimble web search; Serper/SerpAPI/SearchAPI paper search; OpenSearch and Azure AI Search knowledge backends
Sandboxed code execution via Modal and OpenShell with durable artifact capture stored in SQL or S3-compatible storage
Built-in evaluation harnesses for DeepResearch Bench and FreshQA benchmarks with RACE and FACT metrics
Production-ready deployment with OAuth-protected per-user MCP sources, NeMo Guardrails middleware, optional content encryption, and OTEL tracing
Docker Compose and Helm deployment options with published NGC containers for backend, frontend, and release chart

Use Cases

  • 01Enterprise research automation with governed access to internal knowledge bases and external academic/web sources
  • 02Citation-backed business intelligence reports combining company data with market research and industry papers
  • 03Multi-phase investigation workflows requiring structured planning, concurrent source queries, and synthesis into comprehensive outputs
  • 04AI assistant integration via MCP protocol to expose research capabilities to Claude Desktop, Cursor, and other MCP clients
  • 05Academic literature review automation searching Google Scholar and synthesizing findings with proper citations
  • 06Benchmarking and evaluating custom agentic research workflows against standardized datasets

Related MCP Servers

View more

markitdown

Python tool for converting files and office documents to Markdown.

MCP Server187.8k

firecrawl

The context API to search, scrape, and interact with the web at scale. 🔥

MCP Server187.2k

prompts.chat

f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.

MCP Server171.7k

langflow

Langflow is a powerful tool for building and deploying AI-powered agents and workflows.

MCP Server155.4k

aiq — FAQ

What is the NVIDIA AI-Q Blueprint MCP server?+

AI-Q is an MCP server that exposes enterprise research capabilities through submit/poll/report operations, enabling AI assistants to run citation-backed shallow or deep research workflows. It connects to NVIDIA NIM models, web/paper search APIs, and knowledge layers to deliver governed research outputs.

How do I install and run the AI-Q MCP server?+

Clone the repository, install dependencies with the setup script, configure API keys in deploy/.env, then run 'uv run --project mcp --frozen aiq-mcp-server' with NVIDIA_API_KEY, TAVILY_API_KEY, and AIQ_CHECKPOINT_DB environment variables set. The MCP endpoint defaults to http://localhost:9001/mcp.

Which MCP clients work with AI-Q?+

AI-Q uses the Streamable HTTP MCP protocol and works with any MCP-compatible client supporting that transport. The server intentionally has no authentication and must be deployed on a trusted network or behind appropriate access controls.

What API keys and prerequisites are required?+

NVIDIA_API_KEY from build.nvidia.com is required for model inference. At least one data source is needed: TAVILY_API_KEY, EXA_API_KEY, NIMBLE_API_KEY, YDC_API_KEY (You.com), or SERPER_API_KEY/SERPAPI_API_KEY/SEARCHAPI_API_KEY for paper search. Python 3.11-3.13 and uv package manager are required; no GPU needed when using NVIDIA API Catalog.

Is the NVIDIA AI-Q Blueprint free to use?+

The AI-Q software is open source under Apache License 2.0. However, it requires NVIDIA API keys for model inference and third-party API keys (Tavily, Serper, etc.) for data sources, which may have their own pricing. Self-hosting models via NVIDIA NIM requires appropriate GPU hardware.

Can I use AI-Q without the MCP server?+

Yes, AI-Q offers multiple interfaces: standalone MCP server, REST API with async jobs, command-line interface, web UI, Jupyter notebooks, and direct integration via portable agent skills. Choose the deployment mode that fits your use case.

How do I install aiq?+

Open the source repository on GitHub and follow its README. aiq is a mcp server — MCP Agents Market links you directly to the official repo.

Is aiq free?+

aiq is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches

Research MCP ServersKnowledge & Memory MCP ServersDeveloper Tools MCP ServersBest SkillsBest PluginsBest AgentsTop rated listingsClaude Code toolsBrowse all categories