TTS Model Evaluation Votes and Feedback

A SQLite database contains user votes and feedback from TTS Arena, a platform for comparing text-to-speech models. The dataset was created by Pendrokar and last updated in April 2026. It is designed to help developers identify model faults through community evaluation.

Use Cases

• Analyze user vote patterns to identify which TTS model receives the highest preference scores.
• Evaluate faults of a specific TTS model by querying all votes and comments mentioning that model.
• Compare model performance across different demographic or scenario splits using vote metadata.
• Train a model to predict user preference scores from textual feedback comments.

Strengths

• Data is structured in a SQLite database for efficient querying.
• Last updated in April 2026, indicating recent activity.

Limitations

• Exact row count, column names, and sample size are unknown.
• Contains no actual audio samples, only evaluation votes and feedback.
• Potential for unsafe or unwanted content in user-generated text.

Provenance

Source

HuggingFace Spaces platform via TTS Arena.

Collection Method

Collected from user votes and feedback in an uncontrolled, crowd-sourced evaluation space.

Freshness

Updated April 2026.