As businesses increasingly serve customers across multiple languages, the demand for reliable real-time translation tools continues to grow. Voice AI company Krisp is looking to address that need with the launch of Voice Translation v3, the latest version of its translation platform.
The company has also introduced a self-service Voice Translation API, allowing developers to add real-time voice translation capabilities directly into their own applications.
Krisp said the platform is built for industries where communication accuracy matters, including healthcare, financial services, customer support, and business process outsourcing (BPO). The translation engine supports 61 languages and can convert conversations in real time, even in situations involving background noise, strong accents, or technical terminology.
The new version brings several additions aimed at enterprise users. These include live bilingual transcripts, automatic language detection, custom vocabulary settings, translation quality monitoring, and auditing tools that help organizations review conversations from both customer and agent viewpoints.
Davit Baghdasaryan, CEO and Co-Founder of Krisp, said the company focused on creating a system that performs well in situations where translation mistakes can lead to serious issues.
He noted that many real-time translation solutions have yet to prove themselves in high-stakes environments. Krisp, he said, designed its technology with sectors such as healthcare, insurance, and financial services in mind, where precision is especially important.
Alongside the enterprise platform, the newly launched Voice Translation API gives developers access to the company’s translation technology through a single connection. Applications can send speech input and receive translated voice and text output in return. JavaScript and Python SDKs are available from day one, with additional support expected later.
Robert Schoenfield, Executive Vice President of Licensing and Partnerships at Krisp, said teams building products for telehealth, fintech, and customer support need translation tools that can handle real-world conversations consistently and accurately.
The company also shared results from an existing healthcare deployment. According to Krisp, the system managed 90% of multilingual calls without the help of a human interpreter and delivered 96% translation accuracy across more than eight languages.
Also Read: TCS Launches Rapid Outcome AI Platform With NVIDIA to Accelerate Enterprise AI Adoption
