OpenAI Is Building a New AI Tool That Makes Music from Text

OpenAI is working on a new AI tool that can create full songs from simple text prompts or even add music to existing audio clips.

Imagine typing “upbeat pop song about summer” or humming a melody into your phone, and getting a complete track with instruments, vocals, and rhythm in seconds.

The Information first reported the project, saying the tool could layer background music onto videos or generate guitar riffs to match a singer’s voice.

It’s not clear yet if it will be a standalone app or part of ChatGPT or Sora, OpenAI’s video generator. No release date has been shared.

To make the music sound more human, OpenAI is teaming up with students at the Juilliard School.

They’re helping label thousands of musical scores to teach the AI about things like tempo, harmony, and style. The goal is to go beyond random notes and create music that actually feels expressive.

• OpenAI Developing AI Music Tool That Turns Text Prompts into Full Songs
• OpenAI’s New Project Aims to Create Human-Like Music from Simple Text and Audio Inputs
• OpenAI Partners with Juilliard Students to Build Next-Gen AI Music Generator

Other companies are already in the game. Google has MusicFX, and startups like Suno let users turn text into songs. Some of those tracks have even gone viral. But with OpenAI’s massive user base and tech power, this new tool could make high-quality music creation available to anyone—filmmakers, podcasters, or just people messing around at home.

Not everyone is excited. Musicians worry about losing jobs, and there are growing legal fights over whether AI companies can train on copyrighted songs without permission.

Also Read: Jungkook's "Seven" Sets Record as Longest-Charting Asian Solo

OpenAI hasn’t commented officially, but the project shows the company is still pushing into creative tools. If it works, this could change how music is made—not just for pros, but for everyone.