AI & ML MCP Servers

Servers for integrating AI services directly into your agent workflows. Image generation, speech-to-text, embeddings, model APIs, and vector stores all live here. Useful when you're building agents that chain multiple AI capabilities together or need to call specialized models for specific subtasks.

832 servers in this category · Showing 1–48

Frequently Asked Questions about AI & ML MCP Servers

There are MCP servers for OpenAI, Hugging Face, vector databases like Pinecone and Weaviate, and various embedding services. They let agents orchestrate multi-model workflows.

Yes — AI/ML MCP servers enable agents to call external models, store and query embeddings, and pipe outputs between services, all from a single conversation.

Several servers integrate with Ollama, llama.cpp, and other local inference tools, so your agent can leverage on-device models without sending data to external APIs.