WavespeedMCP is a Model Control Protocol (MCP) server implementation for WaveSpeed AI services. It exposes WaveSpeed's image and video generation capabilities through a standardized MCP interface, enabling programmatic control of generation pipelines from clients and tools. The server offers advanced image generation, including text-to-image with support for image-to-image, inpainting, and LoRA models, along with dynamic video generation that can transform static images into videos with configurable motion. It emphasizes robust, production-ready operation with optimized API polling, intelligent retry logic, and detailed progress tracking. Resource handling is flexible, supporting outputs as URLs, Base64-encoded data, or local files. Errors are managed through a dedicated exception hierarchy, enabling precise identification and recovery. Logging is robust and configurable, with console or rotating file outputs, and adjustable log levels. Configuration is highly flexible, supporting environment variables, command-line arguments, and JSON configuration files, plus per-request timeouts and a separate total wait timeout for polling results. The project follows a modular architecture with server, client, utilities, and exception modules and provides installation, configuration, and usage guidance, including Claude Desktop integration.