Log inSubmit server
Speak AI logo

Speak AI

Search, upload, transcribe, analyze, and manage Speak AI media through Claude, ChatGPT, and other MCP clients.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Speak AI

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "speak-ai": {
      "command": "npx",
      "args": [
        "-y",
        "@speakai/mcp-server"
      ],
      "env": {
        "SPEAKAI_KEY": "YOUR_VALUE_HERE",
        "SPEAK_API_KEY": "YOUR_VALUE_HERE"
      }
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Values are written into the snippet above so you can paste it as is. They never leave your browser.

Automated check passed— started and listed 189 tools correctly (1d ago).

Overview

The Speak AI MCP server connects MCP-compatible assistants to Speak AI workspaces for media uploads, transcript access, AI insights, and workspace management. It supports remote HTTP access at Speak AI's MCP endpoint, with OAuth or bearer-token authentication, plus local-file uploads through stdio and CLI usage. Reach for it when an agent needs to search recordings, inspect transcripts, summarize meetings, ask questions across media, or automate media organization.

Use cases

  • Search customer interviews for recurring themes
  • Summarize meetings into decisions and action items
  • Extract transcript quotes with speaker context
  • Upload and analyze recordings from video URLs
  • Organize and update media files in a workspace

Key features

  • Media upload from URLs and local files
  • Transcript and caption retrieval
  • AI summaries, topics, sentiment, and action items
  • Workspace media search and management
  • AI Chat over media and folders
  • Bulk upload and speaker updates

Tools189

What Speak AI lets your agent do. Expand a tool for its parameters.

Verified live~58.6k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 1d ago

Recent health check succeeded.

7/7 checks healthy over the last 39d

What it can do

From the 189 tools this server returned when we ran it; 189 declare their own safety hints.

Runs commands · 1Can delete or modify data · 89Writes files · 4Handles credentials · 9Makes network requests · 63Read-only tools · 99
Network at startup3 connections
Files written at startupNone
Install scriptsprepare
npm provenanceNot published
Package age211 days
Maintainers1

Observed 2026-10-21 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 9d ago via OSV.dev · @speakai/mcp-server (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

npm downloads4.2kPackage downloads in the last 30 days.
Last commit2d agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed189Callable tools this server registers over MCP.
Directory activity1 viewsConfig copies, upvotes, and views on AllMCPs.

How Speak AI works

What Speak AI MCP server does

The Speak AI MCP server exposes Speak AI media and AI Chat capabilities to MCP-compatible assistants. It can search a workspace, upload media, check processing progress, retrieve transcripts and captions, and request AI-generated topics, sentiment, summaries, and action items.

It also supports media administration. Agents can edit transcript text, rename speakers, update metadata, move files, mark favorites, reanalyze media, or permanently delete files. AI Chat tools can ask questions about media, folders, or the full workspace, then retrieve conversation history, rename chats, submit feedback, and manage favorite prompts.

The service is suited to interviews, sales calls, research sessions, webinars, podcasts, and team meetings stored in Speak AI.

How it works

The Speak AI MCP server is available through the remote endpoint https://api.speakai.co/v1/mcp. Clients can authorize with OAuth through a permission flow or send an API key as a bearer token. Once connected, the assistant invokes MCP tools against the user's Speak AI workspace rather than importing the recordings into the client.

Media processing is asynchronous. Upload operations return a media identifier, and the client should check get_media_status until processing completes before requesting insights or a transcript. upload_and_analyze_batch accepts up to 25 URLs and processes five at a time, reporting individual failures instead of stopping the entire batch.

Supported URL uploads include direct or public files, pre-signed S3 URLs, and links from many social and video services. Local file upload is available only through stdio and CLI deployments, not the remote server.

Setup and configuration

For Claude.ai and Claude Desktop, add the custom connector URL and authorize access in the browser. ChatGPT setup uses Developer Mode, a custom app connection, OAuth, and per-chat connector activation. Cursor and VS Code provide installation links that register the HTTP endpoint and open the authorization flow.

Claude Code can install the Speak AI plugin from the official Claude Code marketplace or add the endpoint manually with HTTP transport. Manual configuration uses an Authorization header with Bearer <your-key>. The documented shell variable for the Claude Code command is SPEAKAI_KEY; the OpenClaw setup uses SPEAK_API_KEY.

API keys are generated from the Speak AI developer settings. OAuth avoids manually placing a key in the client configuration.

Tools and capabilities

  • Upload media from URLs, local files, or batches of URLs.
  • List and filter workspace media with pagination and optional transcript-related data.
  • Retrieve processing status, transcripts, captions, and AI insights.
  • Search media and ask AI questions about recordings, folders, or the workspace.
  • Edit transcripts, rename speakers, update metadata, move files, favorite items, and delete media.
  • Review AI Chat history, messages, prompts, usage statistics, and response feedback.

The upload flow accepts direct file URLs and supported social or video page links, including YouTube, TikTok, Instagram, X/Twitter, Facebook, Reddit, SoundCloud, Twitch, Dailymotion, Streamable, Snapchat, Pinterest, Tumblr, Bilibili, VK, OK.ru, and Rutube.

Limitations and notes

Processing is not immediate: callers should poll get_media_status before requesting downstream analysis. Batch uploads are limited to 25 URLs per call and five concurrent uploads. Bulk speaker renaming supports up to 500 media files in one call.

Local file upload does not work through the remote HTTP server. It is restricted to stdio and CLI use. ChatGPT requires Developer Mode while the connector is not listed in its app store, and Speak AI must be enabled for each chat where it is needed.

API access and AI analysis may depend on the Speak AI account and the applicable analysis cost, which can be checked with get_analysis_quote.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Speak AI

Speak AI is an MCP server that connects Claude, ChatGPT, Claude Code, Cursor, and VS Code to Speak AI workspaces. Its main tools upload and search media, retrieve transcripts and captions, generate AI insights, ask questions about recordings, and manage media metadata, folders, favorites, and processing status.

Install it by adding `https://api.speakai.co/v1/mcp` as a custom connector or MCP endpoint, then authorize with OAuth or configure a bearer API key. Claude Code also supports the official plugin marketplace or a manual HTTP command.

Yes, Speak AI works with Cursor through its provided MCP installation link. Cursor registers the remote endpoint and opens a Speak AI permission flow; you then sign in and allow access.

Yes, but local-file uploads are limited to stdio and CLI deployments. The remote HTTP server supports URL-based uploads, including direct files, pre-signed S3 URLs, and supported social or video page links.

Upload the recording, poll `get_media_status` until it is processed, and then call `get_media_insights` or `get_transcript`. `upload_and_analyze` returns the media ID immediately, while the batch variant handles up to 25 URLs.

Related MCP servers

All in Speech-to-Text All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewSpeak AI AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/speak-ai?style=directory)](https://allmcps.com/mcp/speak-ai)
HTML Embed
<a href="https://allmcps.com/mcp/speak-ai"><img src="https://allmcps.com/api/badge/speak-ai?style=directory" alt="Speak AI on AllMCPs" /></a>