Skip to content
FluxCDNFluxCDN
esc

hexgrad: Kokoro 82M

Text to speech

Kokoro 82M is a lightweight, open-weight text-to-speech model from hexgrad.

Price
357 Toman per 1,000 characters
Endpoint
POST /v1/audio/speech
Published
Apr 23, 2026
Offers
54 voices
Offers
MP3 · WAV
Maker
hexgrad

Free plan prices, in Toman at today's rate. Plans and prices

Price on each plan

Per character of the text read, in Toman at today's rate; a text of 1,000 characters is about a minute and a half of speech.

PlanPrice
Free357 Toman per 1,000 characters
Pro342 Toman per 1,000 characters
Max328 Toman per 1,000 characters

Voices

    54 voices in all: any of them may be asked for by name in the API.

    Hear all voices

    Call it

    curl https://ai.fluxcdn.cloud/api/v1/audio/speech -H "Authorization: Bearer $FLUXCDN_API_KEY" \
      -H "Content-Type: application/json" \
      -d '{"model": "hexgrad/kokoro-82m", "input": "سلام! به FluxCDN خوش آمدید.", "voice": "af_alloy", "response_format": "mp3"}' \
      --output speech.mp3