Data-Driven Analysis of Text-Conditioned AI-Generated Music: A Case Study with Suno and Udio

2025-09-15

AI Summary

This study investigates how users engage in creative practices on text-to-music AI platforms (e.g., Suno, Udio), focusing on linguistic preferences and thematic patterns in prompts, tags, and lyrics. Drawing on large-scale, real-world user data collected between May and October 2024, it presents the first systematic, data-driven analysis of such platforms. Methodologically, the work employs state-of-the-art text embeddings, dimensionality reduction, clustering, automated annotation, and interactive visualization to identify recurrent creative themes—such as nostalgia and affective storytelling—as well as pervasive multilingual code-mixing. Its principal contribution is the discovery and formalization of a “meta-tag guidance” mechanism: users explicitly modulate stylistic, structural, and affective dimensions via structured lexical prefixes (e.g., “[80s synthwave]”). The study releases open-source code and datasets, establishing an empirical foundation and methodological framework for AI musicology, human-AI co-creation research, and computational studies of cultural practice.

Technology Category

• Humans and AI: Crowd Sourcing and Human Computation
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Online AI platforms for creating music from text prompts (AI music), such as Suno and Udio, are now being used by hundreds of thousands of users. Some AI music is appearing in advertising, and even charting, in multiple countries. How are these platforms being used? What subjects are inspiring their users? This article answers these questions for Suno and Udio using a large collection of songs generated by users of these platforms from May to October 2024. Using a combination of state-of-the-art text embedding models, dimensionality reduction and clustering methods, we analyze the prompts, tags and lyrics, and automatically annotate and display the processed data in interactive plots. Our results reveal prominent themes in lyrics, language preference, prompting strategies, as well as peculiar attempts at steering models through the use of metatags. To promote the musicological study of the developing cultural practice of AI-generated music we share our code and resources.

Problem

• Analyzing user prompts and themes in AI-generated music
• Investigating prompting strategies and metatag usage patterns
• Exploring language preferences and cultural themes in lyrics

Innovation

• Text embedding models for prompt analysis
• Dimensionality reduction and clustering methods
• Interactive plots for data visualization