Skip to main content
hexgradMar 19, 2025

Kokoro Text to Speech

tts-kokoroPrivateOpen weights

Kokoro is hexgrad’s open-weight 82M-parameter text-to-speech model. It speaks in 54 voices across several languages and returns MP3, Opus, AAC, FLAC, WAV or PCM audio.

$3.50
per 1M characters
$0.1733
per hour of speech
54
voices
54 voices

Names are case-sensitive. Pick one to use it in the request below.

Capabilities

Formatsmp3 · opus · aac · flac · wav · pcm, mp3 by default
Streamingstreaming: true
OpenAI-compatibleclient.audio.speech.create()

Pricing

Billed per 1M characters of input, in USD.

Input$3.50per 1M characters
Speech, about$0.0029per minute
Speech, about$0.1733per hour

Estimated at 825 characters a minute, about 150 spoken words.

Estimate a script

121 minof speech, estimated
$0.35to speak them

Against other speech models

How it comparesamong 15 speech models
Pricecheaper than 100%$3.50
Voicesmore than 100%54

Further right is better. Hover a dot for its model.

Specs

Model ID
tts-kokoro
Endpoint
POST /api/v1/audio/speech
Voices
54
Formats
mp3, opus, aac, flac, wav, pcm · default mp3
Input
Up to 4,096 characters per request
Streaming
Yes
Voice cloning
No
Privacy
Private
Added
Mar 19, 2025

Compare

Every text-to-speech model on Venice, cheapest first.

ModelVoicesFormatsPrivacyPer 1M characters
Kokoro Text to Speech54mp3, opus, aac, flac, wav, pcmPrivate$3.50
Inworld TTS-1.5 Max14wavAnonymized$12.50
xAI TTS v126mp3, wav, pcmPrivate$18.75
ElevenLabs TTS v4 Turbo20mp3Anonymized$46.00
Gradium TTS12wav, pcm, opusAnonymized$47.50
Chatterbox HD (Resemble AI)9wavPrivate$50.00
ElevenLabs Turbo v2.521mp3Anonymized$62.50
Orpheus TTS8wavPrivate$62.50
Qwen 3 TTS 0.6B9mp3Private$87.50
ElevenLabs TTS v420mp3Anonymized$92.00
Qwen 3 TTS 1.7B9mp3Private$113
ElevenLabs Multilingual v220mp3Anonymized$115
ElevenLabs TTS v320mp3Anonymized$115
MiniMax Speech-02 HD15mp3, pcm, flacAnonymized$125
Gemini 3.1 Flash TTS30mp3, opus, wavAnonymized$188

Quickstart

OpenAI-compatible: point any OpenAI SDK at the Venice base URL with a Venice API key.

curl https://api.venice.ai/api/v1/audio/speech \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "tts-kokoro",
"input": "Private AI for everyone.",
"voice": "af_alloy",
"response_format": "mp3"
}' --output speech.mp3
Base URL https://api.venice.ai/api/v1

Parameters

ParameterValuesNotes
modeltts-kokoroRequired
inputUp to 4,096 charactersRequired
voice54 voicesThe model’s own by default
response_formatmp3 | opus | aac | flac | wav | pcmDefault mp3
speed0.25 – 4.0Some models clamp it or ignore it
streamingtrue | falseStreams the audio as it is generated