How the Suno system actually works, from a workflow that ships real tracks. One concept per article: what the model reads, what it ignores, and which parts of a prompt are suggestions versus contracts. Every claim is tested against a real generation from this catalogue's production files, with the exact blocks quoted. Articles publish as their tests are run, in production order rather than syllabus order.

Articles8
Published1
Edition2026
Languageen-GB

The articles.

In series order
Article 01
What Suno reads and what it ignores
Which tag types have the most impact and which are decorative.
TBD
Upcoming
Article 02
How tag order changes generation
First tags are weighted higher — provable with identical tags reordered.
TBD
Upcoming
Article 03
Genre tags vs texture tags vs rhythm tags
Three categories, three different functions, one test.
TBD
Upcoming
Article 04
Why short tags outperform long descriptions
Suno parses concepts, not sentences — shown with a direct comparison.
TBD
Upcoming
Article 05
BPM and key — when they work, when they don't
BPM and key are suggestions to a vocal generation and contracts to an instrumental one. The two-block discipline that makes the grid hold, from a real workflow.
7 min read
Published
Article 06
The difference between style tags and section tags
How [Chorus] and [Drop] function differently from style-field input.
TBD
Upcoming
Article 07
How Suno handles contradiction in tag sets
Why dark + playful works, and what happens when you push it too far.
TBD
Upcoming
Article 08
Extending vs regenerating — when to do which
A decision framework with examples of both going right and wrong.
TBD
Upcoming