Kukedlc/suno-ai-music-dataset

Contents & fields

The dataset contains MP3 audio files, JPG cover art, and a master CSV file (biblioteca_master.csv) with 32 columns of metadata. All metadata is keyed by Suno's UUID. Fields include:

• id——Suno-issued UUID, primary key
• title——Track title
• hoja_id_pipeline——Internal pipeline identifier
• major_model_version——Suno model family (e.g., v5.5)
• model_name——Specific model variant (e.g., chirp-fenix)
• status——Suno processing status
• duration——Track duration in seconds
• is_instrumental——Boolean for instrumental tracks
• tags——Free-form tag string at generation
• tax_radio_destino——Target radio/listener context label
• tax_suno_capability——Suno capability classification
• prompt——Exact text prompt used for generation

Research uses

This dataset is suitable for fine-grained audio classification, text-to-music conditioning experiments, music recommendation and similarity research, benchmarks for music tagging and auto-DJ systems, studies of AI music generator capabilities and failure modes, and training material for derivative music-generation models (subject to license terms).

This card was drafted from the source page; institution, coverage, time span, scale, fields and license are subject to the official page (pending human review).