All Products
Search
Document Center

Intelligent Speech Interaction:Pricing

Last Updated:Aug 28, 2024

This topic describes the billing items, billing methods, and prices of Intelligent Speech Interaction. This topic also describes how to view the consumption details during your use of Intelligent Speech Interaction.

Note

Intelligent Speech Interaction uses the following billing rules:

  • If you adopt the pay-as-you-go billing method, you are charged on a daily basis based on the standard billing rules of Alibaba Cloud. Your consumption on the current day is billed on the next day. Therefore, you must keep your account balance sufficient to avoid overdue payments that may affect service use.

  • The free trial allows you to use only specified Intelligent Speech Interaction services with a limited duration. To improve the processing capability of a speech service, we recommend that you upgrade the service to Commercial Edition or purchase a resource plan for the service based on your business requirements.

    Important

    If Alibaba Cloud finds any user repeatedly uses the free trial with multiple Alibaba Cloud accounts, Alibaba Cloud is entitled to take appropriate measures without notifying the user in advance. The user shall be responsible for the consequences that arise from the violation.

Billing items

The billing system bills you based on the service usage for processing speech or text data and the usage of additional features or resources of Intelligent Speech Interaction.

  • Processing of speech or text data

    Service

    Billing item

    Description

    Real-time speech recognition

    Total duration of the processed speeches

    You can upgrade the service to Commercial Edition and use the service in pay-as-you-go mode.

    Short sentence recognition

    Number of times the service is called

    You can upgrade the service to Commercial Edition and use the service in pay-as-you-go mode.

    Recording file recognition

    Total duration of the processed recording files

    You can upgrade the service to Commercial Edition and use the service in pay-as-you-go mode.

    Speech synthesis

    Number of times the service is called

    You can upgrade the service to Commercial Edition and use the service in pay-as-you-go mode.

  • Additional features or resources

    Feature or resource

    Description

    Self-learning platform

    To improve the recognition accuracy of the words and sentences specific to your business field, you can create custom linguistic models on the self-learning platform.

    Speech synthesis speaker customization

    Based on the deep learning technology, the speech synthesis speaker customization service allows you to customize speakers by using a small amount of training data.

Billing methods and prices

Pay-as-you-go

Usage is settled on a daily basis. Tiered pricing is offered as the call volume increases. The greater the volume, the greater the discount rate.

Service

Standard unit price

Daily usage

Discounted unit price

Real-time speech recognition

USD 1.40 per hour

0‒299 hours

USD 1.40 per hour

USD 1.40 per hour

300‒999 hours

USD 1.20 per hour

USD 1.40 per hour

1,000‒2,999 hours

USD 1.00 per hour

USD 1.40 per hour

3,000‒4,999 hours

USD 0.86 per hour

USD 1.40 per hour

≥ 5,000 hours

USD 0.70 per hour

Short sentence recognition

USD 1.40 per thousand calls

0‒299 thousand calls

USD 1.40 per thousand calls

USD 1.40 per thousand calls

300‒999 thousand calls

USD 1.20 per thousand calls

USD 1.40 per thousand calls

1,000‒2,999 thousand calls

USD 1.00 per thousand calls

USD 1.40 per thousand calls

3,000‒4,999 thousand calls

USD 0.86 per thousand calls

USD 1.40 per thousand calls

≥ 5,000 thousand calls

USD 0.70 per thousand calls

Recording file recognition

USD 1.00 per hour

0‒299 hours

USD 1.00 per hour

USD 1.00 per hour

300‒999 hours

USD 0.95 per hour

USD 1.00 per hour

1,000‒2,999 hours

USD 0.90 per hour

USD 1.00 per hour

3,000‒4,999 hours

USD 0.85 per hour

USD 1.00 per hour

≥ 5,000 hours

USD 0.60 per hour

Speech synthesis

USD 1.40 per thousand calls

0‒299 thousand calls

USD 1.40 per thousand calls

USD 1.40 per thousand calls

300‒999 thousand calls

USD 1.20 per thousand calls

USD 1.40 per thousand calls

1,000‒2,999 thousand calls

USD 1.00 per thousand calls

USD 1.40 per thousand calls

3,000‒4,999 thousand calls

USD 0.86 per thousand calls

USD 1.40 per thousand calls

≥ 5,000 thousand calls

USD 0.70 per thousand calls

Note
  • Free trial

    The following rules on the use of the free trial have been applied since March 1, 2020:

    • The trial is free of charge. The billing policy may be subject to changes. Follow the latest bulletins on Alibaba Cloud international site (alibabacloud.com).

    • You have unlimited usage of the short sentence recognition, real-time speech recognition, and speech synthesis services on each calendar day if the number of concurrent calls does not exceed two. Intelligent Speech Interaction provides you with a three-month free trial for these services.

    • The recording file recognition service can recognize recording files with a maximum total duration of 2 hours on each calendar day. Intelligent Speech Interaction provides you with a three-month free trial for the service.

  • Commercial Edition

    The following rules on the use of Commercial Edition have been applied since June 10, 2019:

    • If you require more than two concurrent calls of recording file recognition or need to recognize recording files longer than 2 hours in length on each calendar day, upgrade the corresponding service to Commercial Edition.

    • Services in Commercial Edition are billed based on your service usage per calendar day. No free quota is available. No fee is charged if you do not use the services. The system automatically calculates and bills your whole daily usage at 24:00 (UTC+8) every day. Bills may be generated with a specific delay.

    • We recommend that you do not downgrade services in Commercial Edition to the free trial. After the downgrade, the services become unavailable because the number of available concurrent calls is decreased to 0.

    • Detailed information:

      • For services billed based on duration, the system accumulates the duration of the speech processed by each service call. The duration is measured in seconds and rounded down. For example, if the speech processed by a service call is 22.8 seconds in length, the duration is recorded as 22 seconds.

      • For services billed based on the number of service calls, failed calls are not counted.

      • All usage on each calendar day is billed at the tiered price for the total usage that is calculated at 24:00 on that day. For example, if the short sentence recognition service is called 500 thousand times in a day, the charge for that day is calculated by using the following formula: 500 × USD 1.20 = USD 600.00. In this example, USD 1.20 is the tiered price that corresponds to the usage of 300 to 999 thousand calls.

      • The number of calls to the speech synthesis service is calculated based on the number of UTF-8 encoded characters passed in each service call. Each Chinese character, English letter, or full-width or half-width punctuation mark is counted as one valid character. If 100 or fewer characters are passed in a service call, one call is billed. If 101 to 200 characters are passed in a service call, two calls are billed. If 201 to 300 characters are passed in a service call, three calls are billed. A maximum of 300 characters can be passed in one service call at a time. For example, two calls are billed for a service call with 102 characters, and three calls are billed for a service call with 201 characters.

Billing details of additional features and resources

  • Self-learning platform

    • After technology upgrades, Alibaba Cloud allows you to create custom linguistic models on the self-learning platform for free.

    • You can create a maximum of 10 custom linguistic models.

    By using custom linguistic models, you can optimize the recognition of speech data in business-specific subfields to improve speech recognition accuracy.

  • Speech synthesis speaker customization

    The price of this service varies with the business scenarios, data volume, speaker types, and possession of intellectual property rights. Therefore, no unified quotation is offered.

    If you need to use this service, contact us at nls_support@service.aliyun.com. The relevant responsible person will follow up and offer a quotation after a detailed discussion with you.

View consumption details

  1. Log on to the Alibaba Cloud international site (alibabacloud.com).

  2. Move the pointer over the profile picture in the upper-right corner and select Billing.

  3. In the left-side navigation pane of the Expenses and Costs page, choose Bills > Bill Details.

  4. Click the Consumption by Bill, Billing Details, and View Usage Details tabs to view the detailed information about your consumption.

    image

Notes on refunds

Fees incurred in pay-as-you-go mode cannot be refunded.