CAMB.AI launches real-time dubbing SDK for 150 languages via NVIDIA

The launch of the CAMB.AI real-time dubbing SDK marks a shift toward decentralized localization, allowing broadcasters to process multilingual feeds without the latency of external cloud hops. By integrating with NVIDIA Holoscan for Media, the solution moves AI-driven audio from a post-production luxury to a standard live production component. This capability is particularly critical for rights holders in fragmented markets like Europe or global sports leagues seeking to maximize reach without massive commentary teams. Watch for adoption rates among Tier 1 sports broadcasters to see if AI dubbing becomes the default for secondary language feeds in 2027.

CAMB.AI's integration with NVIDIA Holoscan for Media places it within a broader ecosystem of GPU-accelerated video processing tools that NVIDIA has been building for broadcast and streaming applications. At IBC 2026, NVIDIA's Holoscan for Media platform served as the underlying infrastructure for multiple real-time AI video workflows, and Nokia and NVIDIA have deepened their partnership with a $1 billion investment from the chipmaker into Nokia's RAN strategy, signaling NVIDIA's aggressive push into telecom and media infrastructure simultaneously. The Holoscan for Media framework allows developers to build low-latency AI pipelines that process video and audio streams on NVIDIA GPUs, a capability that CAMB.AI is using to run its dubbing models directly within broadcaster infrastructure rather than relying on external cloud services. The competitive landscape for AI-driven localization in live broadcasting is intensifying, with Panjaya and NVIDIA debuting live AI dubbing for multilingual broadcasts earlier this month. Ericsson launched its AI in RAN commercial software subscription on June 11, 2026, claiming up to 20% higher downlink throughput and up to 10% better spectral efficiency across more than 15 live deployments, demonstrating that AI-driven automation is moving from pilot to production across the entire media delivery chain. For CAMB.AI, the business model hinges on convincing rights holders and broadcasters that real-time dubbing can replace traditional multi-language commentary teams, which typically cost between $50,000 and $200,000 per match for a single language feed. The Streaming SDK approach allows broadcasters to integrate localization into existing production workflows without re-architecting their playout systems. On the technical side, NVIDIA's Holoscan for Media platform provides the GPU-accelerated pipeline that enables sub-second latency for AI inference on live video and audio streams. Nokia is combining with AWS and Databricks to build a telco AI control layer, claiming automation rates higher than 90 percent and service delivery times of four hours or less, illustrating the broader industry trend toward AI-driven automation at scale across network and media infrastructure. For CAMB.AI's Streaming SDK, the critical benchmark is whether voice cloning and terminology customization can maintain broadcast-quality audio at the frame rates required for live sports, where even 200 milliseconds of additional latency can create noticeable sync issues between commentary and on-screen action. NVIDIA's MARS platform, which CAMB.AI also references, provides the model training and fine-tuning infrastructure that allows broadcasters to customize voice profiles for specific commentators or brand voices across their 150 supported languages.