Leaderboard · Model
OpenAI Speech
Ranked #5 of 11 models by blind A/B votes against a real human recording. Humanness 100 means listeners can no longer tell it from the person.
Standings as of August 31, 2026 · refreshed every 5 minutes
Hear it
“Thanks for calling — I can get that rescheduled for Thursday morning. Does that work for you?”
Background
OpenAI's speech synthesis comes from the same audio stack as its realtime voice products. It is rarely marketed as a standalone TTS product, which makes its blind-test standing here a useful independent measurement.
On the board
Live results — the numbers move as votes land. Latency is measured at the Speko gateway (warm first-audio, n=30); price is the provider's published rate. Method details on the methodology page; raw standings as open data (CC BY 4.0).