transcript1 — Claude Code MCP server: what it does, how to install
Transcribes social media clips and recordings directly in your coding assistant. Delivers plain text, timestamps, and subtitles without leaving the terminal.
claude mcp add --transport http mcp https://api.transcript1.com/mcpWhat transcript1 does
The connector links your coding client with the transcript1 backend via an npx -y @transcript1/mcp stdio runner. It processes links from YouTube, Bluesky, TikTok, Instagram, Facebook, and X. Output comes back with chronological timestamps and speaker separation when available.
Local sound and video tracks travel over a chunked upload pipeline. You can check partial transfer progress and resume if packets drop. Completed jobs convert into plain text, JSON, SRT, or VTT files, with paging support for lengthy speeches.
The bridge stops working without Node.js 18 and a valid TRANSCRIPT1_API_KEY token starting with t1_. You configure this variable inside your client environment so the stdio wrapper can reach the remote endpoint securely.
What it helps with
- extract text and speaker names from a YouTube or X link
- produce SRT or VTT subtitle tracks from online recordings
- send local multimedia files via segmented uploads
- monitor remaining subscription minutes and billing dates
Example prompts for the agent
- Transcribe the video at this Bluesky link and label each speaker
- Fetch the transcript from job 42 in SRT format
- Check my transcript1 account balance and renewal date
How to install transcript1
1Platform CLI: remote server
claude mcp add --transport http mcp https://api.transcript1.com/mcp2Platform CLI
claude mcp add --env TRANSCRIPT1_API_KEY=<TRANSCRIPT1_API_KEY> --transport stdio transcript1 -- npx -y @transcript1/mcpReplace <…> with your own values
1Platform CLI: remote server
codex mcp add transcript1 --url https://api.transcript1.com/mcp2Platform CLI
codex mcp add transcript1 --env TRANSCRIPT1_API_KEY=<TRANSCRIPT1_API_KEY> -- npx -y @transcript1/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"transcript1": {
"url": "https://api.transcript1.com/mcp"
}
}
}{
"mcpServers": {
"transcript1": {
"command": "npx",
"args": [
"-y",
"@transcript1/mcp"
],
"env": {
"TRANSCRIPT1_API_KEY": "<TRANSCRIPT1_API_KEY>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"transcript1": {
"type": "http",
"url": "https://api.transcript1.com/mcp"
}
}
}{
"servers": {
"transcript1": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@transcript1/mcp"
],
"env": {
"TRANSCRIPT1_API_KEY": "<TRANSCRIPT1_API_KEY>"
}
}
}
}Replace <…> with your own values
1Platform CLI: remote server
gemini mcp add -t http transcript1 https://api.transcript1.com/mcp2Platform CLI
gemini mcp add -e TRANSCRIPT1_API_KEY=<TRANSCRIPT1_API_KEY> transcript1 npx -y @transcript1/mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about transcript1
How do I install transcript1 in Claude Code?
Run claude mcp add --env TRANSCRIPT1_API_KEY=<TRANSCRIPT1_API_KEY> --transport stdio transcript1 -- npx -y @transcript1/mcp in your terminal. Ensure Node.js 18 or a later release is already installed.
What does transcript1 do?
It provides tools to convert social videos and uploaded recordings into text. The output includes timestamps, speaker separation, and subtitle exports.
Which output formats does transcript1 support?
Results can be retrieved as raw text, JSON, SRT, or VTT. Long transcripts include pagination so the client receives chunks cleanly.
