Speech Recognition API | Convert Audio to Text Online

Demonstration

长按小程序码进入体验

Why Us

High Recognition Accuracy

The recognition accuracy is over 97% for Chinese Mandarin and over 90% for English.

Multi Languages/Dialects Support

15 languages including Chinese, English, Malay, Indonesian, Bengali, Hindi, Japanese, Korean,Thai, Vietnamese, Russian, German, French, Arabian, and Bulgarian are supported. At the same time, we also provide recognition of various dialects in China.

Various Industries Coverage

Based on continuous training and optimization of models, customized recognition for certain industries like Business, Music Playing, and Video Playing are supported.

Novel Self-learning Platform

It allows you to simply specify hotwords, and upload business-related data to build specific models for better recognition accuracy

Application Scenario

Social Network

By converting short audio messages (< 1 min.) to texts, communication on social network becomes more accurate and convenient.

Games and Recreation

Audio files converted into text messages in games greatly improved users' experience while operating.

Voice Search

ASR allows users to easily and efficiently search the web and access GPS navigation or other services with just their voice.

Human-Machine Interaction

ASR integrates a voice wakeup service so that voice commands sent to terminals initiate operations in real time, improving the interaction between people and machines.

Customers

Goodtech

Goodtech is a software, hardware system integration integrated services company. Its intelligent products integrate the voice recognition capabilities deployed by iFLYTEK on Amazon Cloud（AWS）, which greatly improving user experience and efficiency.

Mediazen

Mediazen is a provider of intelligent voice solutions. By calling iFLYTEK's multilingual speech recognition capabilities deployed on Amazon Cloud (AWS), the availability of intelligent voice solutions and service stability are effectively improved.

Pricing

New User Package

New User Package open to enterprises & individuals

Service calls

100,000

Valid period

3 months

Concurrent channels

Virtual Private Cloud

VPC enables you to use iFLYTEK resources in your own VPC.

VPC

customized

Concurrent channels

Business Package A

Resource package based on service calls

Service calls

1,000,000

Valid period

1 year

Concurrent channels

Business Package B

Resource package based on service calls

Service calls

5,000,000

Valid period

1 year

Concurrent channels

Business Package C

Resource package based on service calls

Service calls

10,000,000

Valid period

1 year

Concurrent channels

Q&A

1. How many languages are currently supported for Short Form ASR?

We supported 15 languages for Short Form ASR, including Chinese, English, Malay, Indonesian, Bengali, Hindi, Japanese, Korean, Thai, Vietnamese, Russian, German, French, Arabian, and Bulgarian.

2. What's the difference between dynamic correction enabled or not enabled?

Dynamic correction not enabled: the recognition result is returned in real-time, and the result returned each time is an addition to the previous result;

Dynamic correction enabled: the real-time return of the recognition result, the result returned each time may be an addition to the previous result, or it may be a correction of a previous return;

3. How many concurrent channels does the Short Form ASR WebAPI support?

WebAPI is a server-side solution that has no limit regarding platforms or languages. It supports 5 concurrent channels by default. If there is a business need to increase the number of concurrent channels, please contact our sales team.

4. What are Service Calls?

Service call means API call. It is the process of a client application submitting a request to an API and that API retrieving the requested data from the server and delivering it back to the client.