MCP Servers

Amazon Bedrock MCP — invoke foundation models (Claude, Llama, Titan, Nova) hosted on AWS Bedrock with guardrails and knowledge bases.

Checks

Metadata only

Evidence

1 source

Anthropic API MCP — access Claude models with extended thinking, tool use, vision, and prompt caching from AI agent workflows.

Checks

Metadata only

Evidence

1 source

Apple Shortcuts MCP — run macOS/iOS Apple Shortcuts automations from AI agents to control native apps and OS features.

Checks

Metadata only

Evidence

1 source

AssemblyAI MCP — transcribe audio, detect topics, summarize calls, and run sentiment analysis via AssemblyAI Speech AI.

Checks

Metadata only

Evidence

1 source

Azure OpenAI MCP — access GPT-4o and o-series models hosted on Azure OpenAI Service with enterprise compliance and private networking.

Checks

Metadata only

Evidence

1 source

Cohere MCP — access Command models for generation, reranking documents, embedding text, and NLP classification tasks.

Checks

Metadata only

Evidence

1 source

Query 8,500+ curated awesome lists with over 1 million items to help agents find best-in-class resources for any topic.

Checks

Metadata only

Evidence

1 source

Deepgram MCP — real-time speech-to-text transcription, speaker diarization, and voice intelligence via Deepgram's ASR API.

Checks

Metadata only

Evidence

1 source

DeepSeek MCP — chat with DeepSeek AI models including reasoning mode, multi-turn sessions, function calling, and cost tracking.

Checks

Metadata only

Evidence

1 source

fal.ai MCP — run serverless AI inference for image generation (FLUX, SD), video synthesis, and audio with fal.ai's GPU cloud.

Checks

Metadata only

Evidence

1 source

Gemini API bridge MCP — access Google Gemini Pro and Flash models through the MCP protocol from any MCP-compatible AI agent.

Checks

Metadata only

Evidence

1 source

Connect an agent to GitHub repositories, issues, pull requests, and code workflows.

Checks

Metadata only

Evidence

1 source

Vertex AI MCP — access Gemini, PaLM, and custom-trained models on Google Cloud Vertex AI with model garden and tuning support.

Checks

Metadata only

Evidence

1 source

Groq MCP — access ultra-fast Groq LLM inference for Llama, Mixtral, and Gemma models with sub-100ms response times.

Checks

Metadata only

Evidence

1 source

Search Hugging Face models, datasets, Spaces, papers, and documentation, then run only the community tools enabled in the user's Hub settings.

Checks

Metadata only

Evidence

2 sources

LangChain MCP adapter — expose LangChain tools, chains, and agents as MCP tools for use with any MCP-compatible AI client.

Checks

Metadata only

Evidence

1 source

Mistral MCP — access Mistral Large, Codestral, and Ministral models for coding, reasoning, and multilingual tasks.

Checks

Metadata only

Evidence

1 source

Ollama MCP bridge — run local Llama, Mistral, Qwen, and other Ollama models through the MCP protocol for offline AI inference.

Checks

Metadata only

Evidence

1 source

OpenAPI MCP — connect any HTTP/REST API using its OpenAPI specification, letting AI agents discover and call arbitrary APIs automatically.

Checks

Metadata only

Evidence

1 source

OpenAPI schema MCP — token-efficient access to OpenAPI/Swagger API specifications via MCP Resources for AI agents building API integrations.

Checks

Metadata only

Evidence

1 source

OpenRouter MCP — access 200+ AI models from Anthropic, OpenAI, Google, Meta, and others via a unified routing API.

Checks

Metadata only

Evidence

1 source

Postman MCP — interact with the Postman public API to manage workspaces, collections, environments, and API definitions from AI agents.

Checks

Metadata only

Evidence

1 source

Replicate MCP — run thousands of open-source AI models for image generation, video, audio, and NLP via Replicate's cloud API.

Checks

Metadata only

Evidence

1 source

Together AI MCP — run 200+ open-source models (Llama, Mistral, Falcon, FLUX) on Together AI's cloud inference infrastructure.

Checks

Metadata only

Evidence

1 source