AI-Media Unveils LEXI Encoders To Modernize Live Broadcasts

AI-Media announced the commercial launch of two new encoder products, the LEXI Text Encoder and LEXI Voice Encoder, at NAB Show 2026, marking the first new encoder hardware release from the company in over a decade. The devices are designed for modern live production, offering 4K broadcast capability via 12G-SDI, expanded processing to run AI-driven features at the edge, and a subscription-aligned commercial model under HaaS.

The product briefing highlights a mix of hardware and software-first upgrades that matter for production and engineering teams. Key features include:

• •4K Broadcast Ready (`12G-SDI`) for single-cable high-bandwidth video workflows
• •Upgraded Processing Power to support low-latency AI tasks and future firmware-model updates
• •`LEXI Live Sync` (formerly CCMatch), a configurable timing alignment system to keep captions and translations synchronized with video
• •Flexible Integration across SDI, IP, and cloud-based workflows, enabling hybrid on-prem and cloud pipelines

The LEXI Voice Encoder specifically targets audio pre-processing for downstream ASR and translation, providing AI sound separation and background noise removal to improve transcription and multilingual voice translation quality. The hardware is positioned to offload front-line cleaning and alignment tasks from cloud services, reducing bandwidth and round-trip latency for live events.

Business and metrics context: This product launch aligns tightly with AI-Media's strategic transition to recurring revenue. The company reported $30.0M annual recurring revenue, an 80% increase year-over-year, and a shift toward higher-margin technology revenue, with group gross margins expanding to 70% in 1H FY26. The encoder rollout is timed with a move to a HaaS commercial model, which management expects will smooth revenue recognition and accelerate predictable ARR growth.

Why it matters: Broadcasters and live-event producers face increasing demand for low-latency, accessible content across linear and OTT platforms, and captioning plus multilingual workflows are now operational must-haves. By embedding AI-driven audio clean-up and caption-sync capabilities into edge encoders, AI-Media reduces dependency on cloud-only stacks and lowers end-to-end latency and operational cost for high-volume live programming. For ML engineers, the platform signals greater emphasis on deployable inference at the edge and lifecycle-managed models distributed via firmware or cloud orchestration.

The move positions AI-Media to compete on infrastructure as much as captioning quality. The combination of hardware upgrades, LEXI feature set, and HaaS means customers trade capital purchases for predictable operating spend, while AI-Media captures ongoing service revenue. This is consistent with peer strategies that monetize models and inference close to the data source to reduce cloud spend and meet strict timing SLAs for live media.

Adoption velocity among major broadcasters and integration with cloud captioning and translation pipelines are near-term indicators to monitor. Watch for developer-facing details about model update semantics, supported codecs and protocols, and latency benchmarks for on-device denoising and real-time ASR preconditioning.

"This marks a significant step forward for AI-Media," said Bill McLaughlin, Chief Product Officer. The company is betting that hardware-enabled AI delivered under a subscription model will convert recent ARR momentum into a durable, higher-margin revenue base.