Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

Text-to-music

Text-to-music is the task of generating music audio from a written description, such as genre, mood, instruments and tempo. Some systems also write and sing lyrics.

Also called: AI music generation, music generation, AI song generator

By Divya Sharma · Checked Nov 15, 2026 · How we review

You describe what you want, for example "slow lo-fi piano with soft rain, 70 bpm", and the model returns an audio clip. Song-focused services such as Suno and Udio can generate full tracks with vocals from a prompt and lyrics.

How it works

Two families dominate. Language-model approaches turn audio into discrete tokens with a neural codec, then predict the tokens one by one, as a text model predicts words. Meta's MusicGen is an open example: a single transformer that works on several streams of audio tokens and can be conditioned on text or a melody. Diffusion approaches, such as Stable Audio 3 Medium, generate audio by denoising, much like a diffusion model for images.

Uses

Background music for videos, podcasts and games, quick demos for songwriters, and variations on a theme. Some tools can extend a clip, separate stems or follow a hummed melody.

Related terms

Text-to-music is a sibling of text-to-speech: both generate audio, but speech models aim for clear words while music models aim for rhythm, harmony and timbre.

Limits

Long pieces can lose structure, and lyrics can come out garbled. Rights are unsettled: training on copyrighted recordings has led to lawsuits, and terms for commercial use of the output differ by service, so read them before publishing.

Related terms

Text-to-speech
Text-to-speech (TTS) is software that turns written text into spoken audio. Modern TTS uses neural networks that predict the sound of speech from the text, then render it as a waveform.
Diffusion model
A diffusion model is a generative model that learns to turn random noise into data, such as an image, by removing a little noise at a time over many steps.
Autoregressive model
An autoregressive model generates a sequence one piece at a time, predicting each next token from all the tokens before it and feeding that output back in. Most large language models work this way.
Voice cloning
Voice cloning is the use of AI to make a text-to-speech system speak in a specific person's voice, learned from a recording of that person. Some systems need only a short sample.
Model license
A model license is the legal terms under which a model's weights, and sometimes its code and data, may be used, changed and shared. Terms range from permissive open licences to non-commercial or custom restricted ones.

Sources