As part of our "Audio Asset Utilization Campaign - 1,000 Hours of AmiVoice API Speech Recognition Free!", we will convert a total of 1,000 hours of your voice data into text for free. Use this campaign in the consideration and verification phase of your business, such as analyzing conversation data at contact centers or launching generative AI projects!

[Campaign content]
AmiVoice APIYou can use this to convert audio data into text for a total of 1,000 hours for free.
*You can continue to use your account even after the campaign ends.
*This applies to the Asynchronous Interface of the general-purpose batch engine. However, if you would like a different engine, please contact us separately.

[Subject]
The following organizations are currently developing or planning to develop products or systems that utilize voice recognition:

  • Companies/Corporate
  • Government agencies and local governments
  • School corporation
  • Medical institution
  • Other organizations

[Application period]
2025 10 / 1 - 2 / 31
*Please note that depending on the application situation, it may take some time for us to respond or we may close applications earlier than planned. We appreciate your understanding in advance.

[How to apply]
Please feel free to contact us.

Contact us

As generative AI rapidly spreads, the ability to accurately transcribe conversations between operators and customers is becoming more important than ever in contact centers. This is because low transcription accuracy undermines the reliability of all subsequent processes, including generative AI analysis, response quality evaluation, compliance checks, and knowledge creation. For this reason, speech recognition has become an essential fundamental technology for utilizing AI in contact centers.

No.1 Contact Center Market Share - AmiVoice Communication Suite

"AmiVoice Communication Suite" is the number one voice recognition solution for contact centers, installed in over 600 companies. It features features such as automatic insertion of punctuation marks, output control of fillers (such as "um"), real-time keyword detection, and flexible integration with generation AI. It also responds to the detailed needs unique to the field, such as deliberately recording fillers to analyze the presence or absence of catchphrases in response quality evaluations. This kind of attentive response to the field is one of the reasons why AmiVoice continues to be chosen.

Strong in proper nouns - ISR Studio with rule grammar

Another strength of AmiVoice is the rule grammar installed in the voicebot service "AmiVoice ISR Studio." Rule grammar is a method of recognizing only expressions that follow predetermined grammar (rules), making it possible to recognize only the model numbers and service names of products handled by a company, to "match" similarly sounding words to ensure correct recognition, and to prevent mistakes in specific patterns such as identity authentication and order acceptance.
This significantly simplifies the analysis process at a later stage, leading to improved accuracy of the voicebot as a whole.
In addition, ISR Studio offers a number of domain-specific engines for addresses, names, etc., and supports specifications that are easy to use in practice, such as returning names in katakana.
Rule grammar and address/name engines are also available in "AmiVoice API Private."

Can also be used as an API - Optimize costs by charging only for speech segments

"AmiVoice API" allows you to directly use the high-precision voice recognition engine used in "AmiVoice Communication Suite."
One of its unique features is that it uses a pay-as-you-go model where you are only charged for the spoken portion of a call. Because there is surprisingly long periods of silence during a call, and in many cases the speaking time only accounts for around 8% of the total, costs tend to be lower than with other companies.
You can choose from multiple engines depending on your needs, it is highly customizable for things like word registration, and it supports integration with major CTI solutions, so you can smoothly incorporate call center voice into your company's systems and existing workflows. Another major benefit is the speedy deployment from PoC to production operation.

The strengths of AmiVoice, optimized for the workplace

A major feature of AmiVoice is that it can be optimized to suit the characteristics of on-site work and technical terminology.
While EtoE (End-to-End) engines are becoming mainstream these days, there are many situations in contact centers where technical terms and proper nouns are frequently used and E2E alone is not enough. "AmiVoice API" offers both EtoE and hybrid engines. The hybrid engine is highly compatible with contact centers, as it allows you to replace language models specialized for your industry or business, flexibly add new keywords (service names, model numbers, technical terms, etc.), and is highly effective in tasks where proper nouns are important, such as knowledge extraction.
The key to optimal operation is to understand the strengths of EtoE and hybrid models and use them appropriately depending on the purpose of the site.

Use API for Free