GPU development demands deep expertise across CUDA, parallel computing, and performance optimization. Developers frequently context-switch between documentation, code examples, and best practices scattered across multiple sources. Traditional code assistants lack the specialized knowledge required for high-performance GPU programming.
This blueprint deploys Nsight Copilot—a self-hosted Model Context Protocol (MCP) server that grounds AI coding assistants and agents in authoritative CUDA documentation. Any MCP-capable client can call its search_cuda_docs tool to retrieve relevant documentation and code samples from a curated CUDA knowledge corpus. Developers can run the server locally without sending prompts or code to an external service.
search_cuda_docs tool at /mcp/cuda-docs performs semantic search over a curated corpus of NVIDIA CUDA documentation and code samples—current through mid 2026, including libraries that post-date most model training cutoffs (such as cuTile).0. One of:
0 with ≥ 24 GB of memory (e.g. H100, GH200).NVIDIA Models
Models
Infrastructure
NVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns here.
Use of the models in this blueprint is governed by the NVIDIA AI Foundation Models Community License.
GOVERNING TERMS: Use of this blueprint is governed by the following:
Nsight Copilot is governed by NVIDIA Technology Access Terms of Use and its CUDA content is governed by the License Agreement for NVIDIA Software Development Kits and CUDA Toolkit Supplement to Software License Agreement for NVIDIA Software Development Kits.
ADDITIONAL INFORMATION: By using Nsight Copilot, NVIDIA may collect aggregate usage statistics and details about the query output for the following purposes: (i) to ensure the service is operating as intended; and (ii) to improve and further develop our products and services.
Use of the llama-nemotron-rerank-1b-v2 NIM container is governed by NVIDIA Software License Agreement and Product Specific Terms for NVIDIA AI Products.
Use of bge-m3 & llama-nemotron-rerank-1b-v2 models is governed by the NVIDIA Open Model License Agreement.
llama-nemotron-rerank-1b-v2 is licensed under Llama 3.2 Community Model License Agreement. Built with Llama.
Deploy a self-hosted Model Context Protocol (MCP) server that grounds AI coding assistants and agents in authoritative CUDA documentation—powered by NVIDIA NIM microservices.