Golazo

Hello everyone.

This article explains how to use speech recognition with "AmiVoice API", a speech recognition API for software development. This document outlines the procedures for using a sample program to display recognition results with your own voice.
*Note: The explanation is based on websites and sample programs as of June 2026.

• Introduction
• 1) Create an AmiVoice API account
• 2) Try running the sample program
• How to develop an app using speech recognition?

Introduction

• The AmiVoice API is available to anyone.All speech recognition engines are available for free.60Please note that usage exceeding the allotted time will incur a charge.

• Prior configuration of credit card or invoice payment is required for use. This is necessary even if you only use the service within the 60-minute free tier. Thank you for your understanding.

• When performing speech recognition, audio data is sent to servers managed by Advanced Media, Inc.. Depending on the plan, you can choose whether or not the sent audio data is stored on the server.*1

• Programming knowledge is not required for this article.

• This article assumes the use of a Windows PC.

• A microphone is required for speech recognition. Please prepare a PC with a built-in microphone or connect a microphone to the PC.

So let's get started.

1. Create an AmiVoice API account

step1.
Please access https://acp.amivoice.com/en/ in a web browser and click on the area enclosed in red in the image to open the AmiVoice API account creation screen.

step2.
Apply through the submission form on the AmiVoice API account creation screen.
On the AmiVoice API account creation screen, enter your email address, check the terms of use and SLA, click the checkbox, and then click "SEND(送信)".

step3.
You will receive an email, click the link in the email body to register.

After a while, you will receive an email at the email address you entered. There will be a URL in the body of the email, so click on it to open the screen below. Enter the required information and click "Next(次へ)".

If you select "Corporate(法人)" as your account type, you can choose between "Credit Card Payment(クレジット払い)" and "Invoice Payment(請求書払い)" as your payment method. If you choose invoice payment, please enter the required information. (This may take one to two business days (or about a week if it's the end of the month) for credit checks, etc.)

step4.
Log in from the login page using the registered Account ID and password.
Login page: https://acp.amivoice.com/mypage/en/login

Once your user registration is complete, please log in using your registered Account ID and password from the login page.

step5.
Click the link in the red circle below to go to the credit card registration screen.

*If you selected "Corporate(法人)" and "Invoice Payment(請求書払い)" when creating your account, you do not need to register a credit card separately, so please proceed to step 7.

The following screen is displayed after logging in. To use the AmiVoice API, credit card registration is required. Please click on the area marked with the red circle below to proceed with the registration.

step6.
Enter and save your credit card information.

Supplement: About 3D Secure
When registering a credit card or making a payment,
you may be asked to enter an authentication code (one-time password) sent via SMS.
This is a verification mechanism called 3D Secure, a security measure to prevent fraudulent use.
Please note that the process may vary depending on the card company, including SMS authentication and authorization via a dedicated app.
Furthermore, if you are using a credit card that requires 3D Secure, and authentication is not completed during registration, your credit card will not be valid, and you will not be able to use the AmiVoice API.

step7.
Please record your Service ID and Service Password from the Connection Information on your MyPage.

Click "Connection Information(接続情報)" from the menu on the left of MyPage, and then record the Service ID and Service password circled in red below on the screen that appears (you will use this later).

Important: Please keep the information on this screen strictly confidential and do not share it with anyone else. If it falls into the wrong hands, your paid speech recognition service may be used by someone other than yourself.

2. Download the sample program.

step1.
Install Git.

This sample program is publicly available on GitHub, and to use it, you will need to download (clone) it to your environment using a tool called Git.

step2.
Open the command prompt.

Pressing Enter selected the desktop as the location.

step3.
Download (clone) the sample program from GitHub.

Clone the sample program from this page.
https://docs.amivoice.com/en/amivoice-api/manual/download-client-library/

Please click the link marked with a red circle.

Copy the URL for cloning the sample program from the "Code" (circled in red).

Return to the command prompt you opened in step 2, type "git clone [copied_URL]", and press Enter.

The process is complete when you see a screen like the one below.

3. Try running the sample program.

step1.
This sample will not work with file://, so you will need to open it using a local web server.
Here, we'll explain the simplest method: using the VS Code "Live Server" extension.
First, download VS Code.

The installation method is explained in an easy-to-understand manner here, so please refer to it.

step2.
Open VS Code and click the icon with four squares in a row in the menu on the left side of the screen. Enter "Live Server" in the search bar at the top of the screen and install Live Server (by Ritwick Dey).

step3.
Click "Open Folder" from "File" in the upper left corner, and open the "amivoice-api-client-library" folder that you downloaded earlier.

step4.
From the "Explorer" menu on the left side of the screen, navigate to "AMIVOICE-API-CLIENT-LIBRARY" → "javascript". Right-click on "wrp.html", the second to last file from the bottom, and click "Open with Live Server".
This will allow you to open the sample program on your local web server.

step5.
Enter the Service ID and Service Password, then click "Get Service Authentication Key(サービス認証キーの取得)".
The following screen will be displayed on your browser, so enter the service ID and service password you recorded earlier in the red circled areas. Then click "Get Service Authentication Key(サービス認証キーの取得)" one step below.

step6.
Click "Start Recording(録音の開始)".
The value will be displayed in the APPKEY field at the bottom of the screen. Once you confirm that it is displayed, click "Start Recording(録音の開始)".

step7.
Allow the use of the microphone.
A screen will appear from your browser requesting permission to use your microphone. Click "Allow(許可)". (Audio recorded from the microphone will be sent to Advanced Media, Inc.'s speech recognition server.)

step8.
Speak into the microphone.
When you speak into the microphone, the system will recognize what you're saying and display the results as shown below.
In the following diagram, the phrase "Good morning(おはようございます)" was spoken.

To the right of where the "Start Recording(録音の開始)" button used to be, you can see that the results of the speech recognition are displayed.

Additionally, information such as the progress of speech recognition will be displayed and updated in real time at the bottom of the screen. We won't go into detail here, but for example, the above screen displays the following:

• -> Lines beginning with A: Results of speech recognition processing
  For more details, click here　A Event Packet
• -> Lines beginning with U: Intermediate results of speech recognition processing
  For more details, click here　U Event Packet 

How to develop an app using speech recognition?

Now, in this instance, we have explained the process from applying for the AmiVoice API usage to actually performing speech recognition using a sample program.

The screens and programs used for this speech recognition demonstration are all included in the sample code you downloaded. From here on, developers will need to use this sample code and the manual as references to program and develop their own applications.
AmiVoice API Manual: https://docs.amivoice.com/en/amivoice-api/manual/getting-started/

This time we used a sample using JavaScript, but we also have samples in other languages ​​such as Java and C#, so we hope you will find a sample that suits the programming language you use.

In this case, we used a speech recognition engine called "Conversation_General-purpose(会話_汎用)", but the AmiVoice API offers various speech recognition engines specialized for different purposes. If you have the opportunity, please try various speech recognition engines and experience the differences firsthand.

AmiVoice is an AI speech recognition technology with over 25 years of experience, specializing in Japanese language and technical terms. It has been widely adopted by many government agencies, medical institutions, call centers, and other organizations to improve operational efficiency.
https://acp.amivoice.com/

We encourage you to take on the challenge of developing applications utilizing the AmiVoice API.

This article is a revised and expanded version of the following article originally published in July 2022, updated in June 2026.
https://acp.amivoice.com/en/blog/2021-03-11-110000/