The human benchmark for speech AI.
Woodstitch runs blinded listening studies on text-to-speech models with paid human raters. Every study states its method, its numbers and what it does not show.
Can listeners distinguish AI narration from human narration?
95% CI 42.2–56.1
no length effect detected
lowest of four levels
Listeners did not classify above chance, longer clips did not detectably help, and the most confident judgments were the least accurate. Acoustically the AI differs most on pause rhythm; listeners’ answers tracked pitch range instead. Read the report →
ElevenLabs v4 vs a fine-tuned open model, on audiobook narration
All six listeners independently preferred ElevenLabs, ranging from 76% to 100% of their own decisive judgments. Every listener heard the same passage from both systems, with sides counterbalanced and audio loudness-matched so neither had a level advantage. We built the losing system. We funded this study and published the result anyway. How it was run →
For AI teams
Testing a release candidate? We run blinded human evaluations on your models, voices and scripts. Private by default — paid work never changes a public result.
Evaluate your modelFor listeners
Get paid to compare short audio clips and pick the one that sounds better. Work whenever you want.
Become a raterHow every study is run — listeners, payment, blinding, statistics, publication rules — is in the platform methodology.