Speko's Arena

Leaderboard · Model

Hume Octave

Ranked #4 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

88
humanness (0–100)
31
live blind rounds
613ms
warm first-audio
$0.15
per minute (list)

Hear it

Thanks for calling — I can get that rescheduled for Thursday morning. Does that work for you?

Background

Hume AI positions Octave as emotionally intelligent TTS — a speech model built on an LLM backbone that interprets what it reads. Via the Speko gateway it currently exposes a male voice only, so it competes in male pairings.

On the board

#3ElevenLabs526ms89
#4Hume Octave613ms88
#5OpenAI Speech695ms86

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).