As digitalization advances, language barriers in global communication still pose a significant challenge. Especially in business, education, healthcare, and international events, the need for instant translation is steadily increasing. In response to this need, Microsoft has introduced the Live Interpreter API. This new feature, offered within the Azure ecosystem, stands out by enabling real-time speech-to-speech translation without the need to manually select a target language.
Microsoft’s New Live Interpreter API
The Live Interpreter API is a service that works on the Azure Speech Translation infrastructure and can translate conversations into different languages in real-time. What makes it particularly noteworthy is its ability to eliminate the need to specify the input language. In other words, if a speaker switches from Turkish to English mid-conversation, the system automatically detects and adapts without any manual intervention. Furthermore, thanks to its low latency, the API delivers performance close to that of a human interpreter.
Key Features of the Live Interpreter API
- Real-Time Translation: Translates conversations simultaneously with low latency.
- Azure Speech SDK Integration: Provides easy-to-use SDKs for languages like C# and Python.
- Broad Language Support: Supports 76 languages and 143 language variants, making it suitable for global use.
- Automatic Language Detection (LID): Automatically detects which language is being spoken during a conversation.
- Personal Voice Support: Translations can be done while preserving the speaker’s voice tone and style. (Requires special permission.)
Difference Between Voice Live API and Live Interpreter API
Within the Microsoft ecosystem, Voice Live API and Live Interpreter API are designed to meet different needs:
- Voice Live API: Designed primarily for virtual assistants and real-time dialogue experiences powered by artificial intelligence.
- Live Interpreter API: Focused on real-time speech-to-speech translation. This API facilitates seamless multilingual conversations across different languages.
Supported Regions and Access
The Live Interpreter API is currently in Public Preview. The preview version is available in select regions:
- East US
- West US 2
- Japan East
- West Europe
- Southeast Asia
- Sweden Central
To use the Personal Voice feature, users must apply to Microsoft for special permission.
Pricing Notes
- For current pricing, it’s helpful to check the Azure AI Speech Pricing page.
- Text-to-Speech (TTS) and Personal Voice services are priced based on character usage.
- Speech Translation is priced by the hour. For instance, translating from one source language to a target language costs about $2.50 per hour.
Use Cases for the Live Interpreter API
- Education: Allows students to follow lessons independently in their own language.
- Healthcare: Facilitates safe communication between doctors and patients who speak different languages.
- Call Centers: Enables smooth communication even when the customer and agent speak different languages.
- International Events: Enables participants to listen to conference sessions or broadcasts in their native language in real time.
Future of the Live Interpreter API
The Live Interpreter API is seen as a turning point in multilingual communication supported by artificial intelligence. Although it’s still in the preview stage, its wide language support, low latency, and Personal Voice integration suggest strong potential for cross-industry adoption.
In the future, the API is expected to be available in more regions, offer more flexible pricing, and expand integration options. Particularly in customer experience, educational technologies, and global content production, it holds the potential to set new standards.
