40 mcps available in the DeepSeek directory
Safely read, write, and manage local files with DeepSeek for code editing, log analysis, and file organization.
Access Hugging Face Hub for DeepSeek model downloads, version management, and model card information.
A chatbot implementation using MCP to interact with tools and LLM providers including DeepSeek. Supports connecting to MCP servers, executing tools with retry, and managing sessions.
MCP-like server using DeepSeek API for terminal integration. Features a web-based chat interface, AI-triggered terminal commands, real-time streaming via SSE, and Docker support.
MCP server providing code generation, completion, and optimization using DeepSeek API with tool chaining support. Features generate_code, complete_code, optimize_code, and execute_chain tools with built-in caching.
Production-grade MCP server written in Go with built-in code review, file management, multi-model support (Chat, Coder, Reasoner), and API account management tools.
TypeScript-based MCP server for DeepSeek API with automatic model fallback (deepseek-reasoner to deepseek-chat), multi-turn conversation support, and customizable parameters.
MCP server connecting Claude Desktop with DeepSeek R1/V3 models, optimized for reasoning tasks with 8192-token context window.
Provides access to DeepSeek-R1's reasoning capabilities for LLMs via MCP, with thinking mode and JSON output support.
Delivers DeepSeek reasoning and chain-of-thought content to MCP-enabled clients. Supports both DeepSeek API and local Ollama server.
MCP server enabling clients to use both DeepSeek Chat and Reasoner models with function calling support for up to 128 tool definitions.
Managed integration connecting DeepSeek API to Claude Code via Composio's tooling layer with remote deployment option.
Manage local Kubernetes clusters (minikube, kind) for deploying and scaling DeepSeek inference services.
Interface with Oobabooga Text Generation WebUI for DeepSeek model interaction with advanced sampling parameters.
Connect to Wolfram Alpha for computational mathematics, plotting, and step-by-step solution verification.
Connect directly to DeepSeek's official API with conversation management, model selection, and streaming support.
Connect to DeepSeek models served via vLLM for high-throughput inference with batching and streaming support.
Interface with DeepSeek models through LM Studio for local inference with GGUF model loading and chat management.
Build LangChain agents powered by DeepSeek with tool use, memory, and chain-of-thought reasoning capabilities.
Convert documents between formats (Markdown, LaTeX, PDF, DOCX) with DeepSeek-powered content transformation.
Perform web searches through Brave Search API with DeepSeek for result analysis and information synthesis.
Manage Docker containers and images locally with DeepSeek for troubleshooting, configuration, and deployment.
Scrape and extract structured data from websites with DeepSeek-powered content parsing and data transformation.
Interact with local and remote git repositories for commit analysis, diff review, and branch management.