# Pricing
### 1. Pricing Overview
MaaS supports multiple model types, and pricing depends on the type of model used.
#### Model Classification for Billing
Currently, there are two main model groups:
* **Token-based models**
* **Image-based models**
***
### 2. General Principles
#### 2.1. oken-based Models
Applies to models such as:
* Chat
* Completion
* Embedding
* Rerank
* Speech-to-text / Text-to-speech *(*if converted into tokens*)*
**Pricing Method**
* Tokens are divided into three types:
* **Input tokens**
* **Cache tokens** *(if applicable)*
* **Output tokens**
* Each token type has **its own unit price**.
* Pricing is calculated **per 1,000,000 tokens**.
Pricing Formula
```
Total Cost =
(Input tokens - Cache tokens / 1,000,000 × Input price)
+ (Cache tokens / 1,000,000 × Cache price)
+ (Output tokens / 1,000,000 × Output price)
```
***
#### 2.2. Image Generation Models (Gen-Image)
Applies to models such as:
* Text-to-image
* Image-to-image
**Pricing Method**
Pricing is based on:
* Number of images generated
* Configuration of each image, such as:
* Resolution (kích thước)
* Quality
* Style / Model version
⇒ Each image configuration has a different price.
***
### 3. Detailed Rules by Account Type
#### 3.1. Prepaid Account
**3.1.1.** Usage Principles
* Users must purchase user credits in advance using the Top-up credits feature.
* For each model usage:
* Generated tokens and/or images will be:
* Calculated based on the corresponding pricing
* Deducted directly from the current user credit quota
* When quota = 0:
* All models will be disabled
* Users must purchase additional quota to continue usage
***
**3.1.2.** User Credit Expiration
* User credits are valid for 1 year
* The validity period is calculated from the most recent top-up date
When user credits expire:
* All models will be disabled
* Users can:
* Recover user credits from the Billing page
After recovery:
* Previous user credit quota will be lost
* The system will only use the newly recovered quota
Recovery Expiration
* If more than 7 days have passed since user credit expiration:
* Users must:
* Reactivate billing
* Purchase additional user credits if they wish to continue using the service
> It is recommended that users estimate usage in advance and proactively top up credits to avoid service interruptions due to exhausted quota or expired credits.
>
> (The Quota Alarm feature is under development and will be updated in the future.)
***
#### 3.2. Postpaid Account
**3.2.1. Usage Principles**
* User credits have unlimited validity
* Users can use models without pre-purchasing quota
* Usage will be:
* Recorded in real time
* Aggregated and billed at the end of each month
***
### 4. Usage Monitoring
To view detailed usage:
* Go to the [**Model Usage**](https://aiplatform.console.vngcloud.vn/model-usage) **page**
* You can:
* Select a specific time range
* View details such as:
* Tokens input / output / cache
* Number of generated images
* Corresponding costs
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.vngcloud.vn/vng-cloud-document/ai-stack/ai-platform/model-as-a-service/pricing.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.