Gladia Launches Solaria: Next-Gen Speech Recognition for Global Call Centres and Voice Platforms

Gladia launches Solaria, a next-generation speech-to-text model offering ultra-low latency (<300 ms), 94% accuracy, and support for over 100 languages. Designed for call centres and voice-first platforms, it ensures real-time multilingual transcription, adapts to industry-specific terms, and operates effectively in noisy environments.

Gladia, a provider of advanced AI transcription and audio intelligence solutions, has announced the launch of its latest automatic speech recognition (ASR) model—Solaria. This next-generation speech-to-text (STT) engine is developed to meet the growing demands of global businesses, particularly within the call centre and voice-first communication sectors.

As businesses worldwide seek scalable solutions to bridge language gaps, Solaria is positioned as an enterprise-grade model that supports real-time, multilingual speech processing at high levels of accuracy and speed. The new model supports more than 100 languages, including native-level transcription in key regional languages such as Tagalog, Bengali, Punjabi, Tamil, Urdu, Persian, and Marathi.

Jean-Louis Quéguiner, Co-Founder and Chief Executive Officer at Gladia, comments on the significance of this launch:

"Speech is the most natural way to connect with the world—for the first time, automated speech recognition is closing the divide, enabling humans and AI to truly speak the same language," he says. "With Solaria, we have made a breakthrough in AI-powered voice technology that unlocks new opportunities for businesses, driving efficiency and delivering more seamless, impactful customer experiences across diverse languages and markets. Solaria is built for next-generation voice platforms ready to lead this transformation on a global scale."

Meeting the demands of global communication

With nearly half (49%) of global executives reporting financial losses due to language barriers, the demand for AI-powered multilingual tools is increasing rapidly. Solaria addresses this challenge by offering robust transcription and voice intelligence features tailored for both AI automation systems and human agents in enterprise environments.

Designed for large-scale deployment, Solaria delivers transcription with ultra-low latency of less than 300 milliseconds—enabling real-time customer interaction without unnatural delays. The model also maintains an industry-leading Word Accuracy Rate (WAR) of 94% across widely used languages such as English, Spanish, and French.

Moreover, Solaria stands apart in its ability to adapt to sector-specific requirements. Businesses in industries like healthcare, finance, and customer service can benefit from its ability to integrate customised vocabulary and extract vital information such as names, emails, numerical data, and addresses from spoken conversations.

Enhanced support for voice-driven platforms

Solaria is engineered to operate in high-noise environments such as call centres, utilising adaptive speech processing techniques to maintain consistent accuracy. The model supports real-time code-switching and translation between all supported languages, enabling global customer service operations to scale without the constraints of language limitations.

Gladia’s platform offers compliance with key international data protection standards including GDPR, HIPAA, and SOC 2, ensuring enterprise-grade data security for users in regulated sectors.

In addition to supporting call centres, Solaria’s capabilities extend to other voice-based applications including virtual assistants, interactive voice response (IVR) systems, and AI-powered voice agents. By integrating Solaria, businesses can significantly improve the reliability of automated and semi-automated customer engagement solutions.

Strategic partnerships and future direction

To support developers integrating speech technology into their applications, Gladia partners with LiveKit—a leading open-source framework for AI voice agents. Through this collaboration, developers can implement real-time multilingual translation with seamless access to Gladia’s API, offering out-of-the-box language capabilities.

Gladia’s growth is underpinned by strong enterprise adoption. More than 700 global clients—including Attention, Circleback, Method Financial, and VEED.IO—use Gladia’s speech AI infrastructure. The company’s services are supported by a flexible GPU hosting system deployable across the U.S., Europe, or other global locations, ensuring full compliance with local data and operational regulations.

Following its $16 million Series A funding round in 2024, the launch of Solaria marks another milestone in Gladia’s evolution toward becoming a full-stack audio intelligence infrastructure provider. The company combines transcription, generative AI, and speech synthesis into a comprehensive platform for real-time voice applications.

About Gladia

Gladia was founded in 2022 by Jean-Louis Quéguiner (formerly of OVHCloud) and Jonathan Soto (formerly of MIT and Sigfox). The company offers state-of-the-art speech AI tools aimed at empowering global businesses to enhance communication through intelligent, real-time audio solutions. 

Headquartered in the U.S. and Europe, Gladia currently serves over 150,000 users and maintains infrastructure that supports reliable, secure, and scalable performance for mission-critical operations.

For enterprises seeking to enhance their customer experience, improve operational efficiency, and scale multilingual support, Gladia’s Solaria presents a compelling solution for the future of voice-driven communication.