MCP servers
122,089 servers from the official registry, GitHub and npm. Filter by category, transport and language.
122,089 servers from the official registry, GitHub and npm. Filter by category, transport and language.
Paste its GitHub repo and we list it after a quick review.
Transcribes audio and video files into subtitles and text using local GPU acceleration across AMD, NVIDIA, Intel, and Apple hardware.
$claude mcp add babelscribe -- uvx babelscribeDiagnoses and fixes CUDA, GPU driver, toolkit, and Python AI library version mismatches, in terminals, notebooks, Docker, and CI.
Provides AI agents with version-pinned NVIDIA CUDA-Q documentation, API references, and runnable code examples using local SQLite full-text search.
$claude mcp add cudaq-docs-mcp -- uvx cudaq-docs-mcpEnforces engineering policies in AI coding agent workflows by providing policy-aware preflight checks, citations, and runtime rule enforcement.
Enables AI agents to discover and utilize your entire machine fleet, including GPUs, models, software, and datasets, via existing SSH connections.
$claude mcp add fleet -- uvx agents-fleetEnables coding agents to discover and utilize your entire machine fleet, including GPUs, models, software, and datasets, via existing SSH connections.
$claude mcp add fleet -- uvx agents-fleetAllows AI agents to inspect, manage, and coordinate shared NVIDIA GPU VRAM usage for Ollama and other model runtimes.
Enables coding agents to profile, rewrite, and verify GPU kernels using Nsight, SASS, and Triton through an autonomous speedrun loop.
Routes AI queries between local GPUs and cloud models using intelligent hardware-adaptive cascading to minimize token costs.