Speech.sh MCP server

Speech.sh is a small helper that lets your AI speak out loud. When your assistant finishes a job, it can say a quick summary instead of you having to read the whole thing. It is handy if you like to hear results while you do something else, or if reading long replies on screen is tiring.

What this MCP server does

You ask your AI to say something out loud, or you tell it to speak a summary when it is done. The AI passes the text to this helper. The helper sends that text to the Groq text-to-speech service, which sends back an audio file. The helper then plays that audio through a player on your computer, like ffplay, mplayer, or VLC. You hear the words, and the AI carries on with whatever else you asked.

Click to zoom

What you can do with it

• Hear a short spoken summary when your AI finishes a long task
• Ask your AI to read a paragraph or a reply out loud
• Pick a voice you like from a few options
• Slow down or speed up the speech
• Get spoken updates while you work on something else
• Save the spoken audio to a file instead of playing it

Try asking your AI

• “Say out loud that the report is ready.”
• “Read me the last answer you gave, please.”
• “When you finish summarising these notes, speak a two sentence summary.”
• “Say good morning in a slow voice.”

What it gives back to you

You get sound. Your computer plays the spoken version of the text through your speakers or headphones. If you asked for a file, you get an audio file saved on your computer. In the chat itself, your AI usually just confirms that it spoke the text.

Before you start

Who it's for

Anyone who likes to hear results instead of reading them, and people who want spoken updates while their AI works on long tasks.