Simple, Transparent Pricing

Pay only for what you use. No monthly fees, no hidden costs.

Avatar API

Use Avatar API to generate avatar appearance image using prompt.

Service Unit Price ( USD ) Details
Generate avatar appearance
per image

$0.05

Generate base image of avatar appearance
Build avatar
per request

$4.00

Avatar can be appeared in images, videos and Interactive mode generated via the platform.

Generative API

Generate realistic image and video with avatars.

Service Unit Price ( USD ) Details
Image generation
per image

$0.07

AI-powered image generation with custom avatars, prompts and styles.
Talking avatar video generation
per second output

$0.01

Generate talking avatar video with image and audio input.

Voice Clone API

Clone your own choices of voice to be used in speech api, or during video and Interactive mode.

Service Unit Price ( USD ) Details
Deep Cloning
per request

$3.00

Clone a voice from a 15 mins sample voice data. Deep clone can produce a voice with all the voice features (i.e. Pronounciation, tone, tempo etc) being very close to the sample.

The process takes about 20 mins.

Speech API

Text-to-speech and Speech-to-text model.

Service Unit Price ( USD ) Details
Text-to-speech - Premade
per second output

$0.0025

Generate realistic speech from text, using voices from our premade library or a custom voice.
Support English, Cantonese, Mandarin, and more in the future release.
Text-to-speech - Custom
per second output

$0.003

Generate realistic speech from text, using voices from our premade library or a custom voice.
Support English, Cantonese, Mandarin, and more in the future release.
Speech-to-text
per second input

$0.0005

Support English, Cantonese, Mandarin, Japanese, Korean, and more in the future release.

Mirako Interactive API

Real-time interactive AI conversations and responses

Service Unit Price ( USD ) Details
Live Interactive Session
per minute

$0.07 [2]

Real-time voice chat with your custom avatar.
- Customizable instruction
- Extend it with agentic tool calling
Remark: [2] Pricing include the usage of supported LLM models during the period.

The usage starts to bill when the interactive model has been fully launched, and stops when the model instance has been terminated. Session is either manually terminated via api, or automatically terminated when no active connection. Note that there is a session timeout period (customizable with a minimum of 5 minutes) before automatical termination kicks in.