AssemblyAI MCP Server - diarized captions, chapters, speaker labels, sentiment

check: npx https://www.assemblyai.com/docs

AssemblyAI MCP server. Speech-to-text with speaker diarization, auto-chapters, sentiment, and caption-ready SRT/VTT output. Best-in-class metadata for caption-driven social videos.

MCP Server

Package: assembly-ai-mcp (official)
Transport: stdio
Command: npx -y assembly-ai-mcp

Authentication

Variable	Required	Description
`ASSEMBLYAI_API_KEY`	Yes	API key from assemblyai.com dashboard

Available Tools

Tool	Description
`transcribe_url`	Transcribe audio at a public URL
`transcribe_file`	Transcribe a local audio file
`submit_transcription`	Async submit, returns transcript id
`get_transcript`	Fetch transcript by id

Configuration Example

toml

[[mcp.servers]]
name = "assemblyai"
type = "stdio"
command = "npx"
args = ["-y", "assembly-ai-mcp"]
timeout_seconds = 120
env = { ASSEMBLYAI_API_KEY = "{{INPUT:ASSEMBLYAI_API_KEY}}" }
tools = []

Notes: Use AssemblyAI for the final captioned upload (chapters, speaker labels). Whisper is fine for quick drafts. Free tier 100 hr/mo.

assemblyai/assemblyai

MCP Server

Authentication

Available Tools

Configuration Example

Links