Gives your AI agent pronunciation scoring, speech-to-text, and text-to-speech through a streamable HTTP MCP server. Connects to Brainiall's Azure-hosted APIs with sub-500ms pronunciation assessment that scores phonemes individually, 12-voice TTS at 24kHz, and compact STT with word timestamps. The MCP exposes 10 tools, 8 resources, and 3 prompts for building language learning workflows. You'd reach for this when building tutoring agents that need to evaluate spoken English, generate audio feedback, or transcribe student responses. Requires an API key from the Brainiall portal. Also available are sibling servers for NLP tasks and image processing if you need toxicity detection or background removal alongside speech features.
claude mcp add --transport http io.github.fasuizu-br-speech-ai https://apim-ai-apis.azure-api.net/mcp/pronunciation/mcp