Skip to main content
Model tokay-kk-v1 (Kazakh) ships with six named voices — three female, three male. Pass one by name; omit voice and you get the default, Aigerim.

The catalog

Aigerim

female · warm, professional — the default. Reception and customer support.

Arman

male · confident, professional — support, IVR, announcements.

Ainur

female · bright, energetic — promos and upbeat greetings.

Aruzhan

female · youthful, clear — support and notifications.

Sanzhar

male · neutral, precise — information, readouts, IVR.

Yerlan

male · calm, warm — explanations and long-form.

Always fetch the live list

The catalog grows and changes — don’t hard-code it. Pull it at runtime (no API key needed).
Each entry has name, gender, style, and best_for — enough to build a voice picker or route by use case. See GET /v1/voices.

Auditioning

Generate a one-line sample per voice to pick by ear:

Next steps

Quickstart

From an API key to your first hello.wav.

Streaming audio

Play a long reply sentence-by-sentence.

Latency & quality

Tune nfe for speed or fidelity.

API reference

Every endpoint, with a live playground.