CAMB.AI has entered a partnership with Broadcom to integrate its edge AI text-to-speech technology into Broadcom’s System-on-Chip (SoC) platforms. The collaboration aims to bring on-device multilingual speech capabilities to home entertainment and connectivity systems, reducing reliance on cloud processing and improving accessibility for users worldwide.
CAMB.AI’s voice emulation models have been optimized for Broadcom’s SoC architecture, which powers a wide range of set-top boxes and broadband gateways. The company has introduced a text-to-speech capability that runs on the neural processing unit (NPU) integrated within Broadcom’s chips. The system converts written text into natural, human-like speech in multiple languages directly on the device.
According to CAMB.AI, this setup delivers low latency, energy efficiency, and improved privacy since all processing happens locally rather than through external servers. Users benefit from faster communication and reduced bandwidth consumption, making the technology suitable for applications such as smart devices, e-learning, and customer service systems.
“Our mission at CAMB.AI has always been to break down language barriers for eight billion people around the world and make communication truly universal,” said Akshat Prakash, Co-Founder and CTO of CAMB.AI. “By partnering with Broadcom, we can deliver this capability to consumers globally in a way that is faster, more private, and more integrated into everyday devices than ever before.”
Edge AI Integration on Broadcom Platforms
The integration uses Broadcom’s neural processing unit to handle AI inference locally, eliminating the need to send voice or text data to cloud servers. This on-device design enables faster response times and ensures data stays on the device, aligning with growing privacy and compliance standards across markets.
Broadcom’s hardware is widely deployed in home entertainment and broadband infrastructure, offering a path for large-scale distribution of CAMB.AI’s text-to-speech models. The collaboration focuses on enabling localized user experiences that combine Broadcom’s high-efficiency silicon architecture with CAMB.AI’s multilingual AI models.
“This collaboration showcases the power of bringing together CAMB.AI’s expertise in real-time multilingual AI and Broadcom’s advanced SoC technology,” said Rich Nelson, Senior Vice President and General Manager of Broadcom’s Broadband Video Group. “We are enabling next-generation user experiences that are both highly intelligent and privacy-first.”
The companies are now assessing the potential to extend this integration to include real-time language translation on-device. If implemented, it would enable translation in over 150 languages, creating a unified communication experience without external connectivity.
Applications Across Entertainment and Accessibility
In home entertainment, the collaboration enables direct localization of content within devices. Broadcasters and streaming platforms could generate audio in multiple languages without needing external servers or long processing times. This makes content more inclusive for global audiences and supports accessibility initiatives by providing real-time speech for visually impaired users.
CAMB.AI’s models maintain the tone and emotion of original speakers, ensuring translated or dubbed audio remains authentic. The technology can also convert content into accessibility-friendly formats on demand, expanding use cases for education, customer engagement, and assistive technologies.
CAMB.AI has previously applied its language models in large-scale deployments across live sports and media. In 2024, it became the first company to deliver live, real-time translations of Major League Soccer games into multiple languages. The company has also worked with IMAX, Comcast NBCUniversal, the Australian Open, and Maple Leaf Sports & Entertainment to enhance content delivery through AI speech and translation tools.
Positioning for On-Device AI Growth
The collaboration reflects a growing industry movement toward on-device AI solutions that reduce dependence on external networks. By processing data within the hardware, companies can improve performance, cut energy use, and comply with data protection laws that restrict cloud-based transfers.
For Broadcom, the partnership builds on its position in the home entertainment and broadband sectors, where its chips power many of the world’s leading consumer devices. Integrating CAMB.AI’s models adds multilingual capabilities to its hardware ecosystem, enabling device manufacturers to deliver localized, privacy-conscious user experiences.
CAMB.AI’s portfolio includes its BOLI translation and MARS speech emulation models, available on AWS Bedrock and Google Vertex AI. Extending these models to Broadcom’s hardware infrastructure bridges software-based AI with device-level integration, creating a framework for scalable, real-time localization across global markets.
CAMB.AI operates from offices in the United States, the UAE, India, and Japan. It works with global partners in media, technology, and sports to develop multilingual communication solutions. The Broadcom collaboration expands its reach to hardware-level implementations, bringing its edge AI technology directly into consumer electronics.
