NAB 2026: AI-Media Launches LEXI Text Encoder and LEXI Voice Encoder

Why It Matters

The new encoders enable AI‑powered, multilingual live production at scale, while the subscription model lowers financial barriers for broadcasters and content creators.

Key Takeaways

• First AI‑Media encoder hardware released in over ten years
• Supports 4K 12G‑SDI and hybrid SDI/IP/cloud workflows
• LEXI Voice adds AI sound separation and noise removal
• New HaaS model eliminates upfront capital expense
• Partners NVIDIA, Grass Valley, AudioShake showcase integration

Pulse Analysis

AI‑Media, a long‑standing provider of live‑production infrastructure, used the NAB Show to signal a strategic pivot toward AI‑centric hardware. After a ten‑year lull in new encoder releases, the company unveiled the LEXI Text and Voice encoders, positioning itself to meet broadcasters’ growing demand for real‑time multilingual services. The move reflects broader industry pressure to automate captioning, translation and audio description, especially as OTT platforms and sports broadcasters chase global audiences. According to research, AI‑driven captioning market is projected to exceed $1.2 billion by 2028, driven by regulatory mandates and consumer expectations for accessibility.

Both encoders support native 4K video over 12G‑SDI and can be deployed across traditional SDI, IP‑based, or cloud‑native pipelines. LEXI Live Sync, formerly known as CCMatch, offers configurable timing alignment for captions and translations, ensuring lip‑sync accuracy in multi‑language streams. The Voice encoder adds AI‑driven sound separation and background‑noise suppression, enabling clean voice extraction for real‑time translation without sacrificing audio fidelity. Latency benchmarks show sub‑second turnaround for translation, making the encoders suitable for live sports and breaking‑news events where timing is critical. These capabilities give production teams a single hardware platform that handles text, voice and analytics in one chassis.

Perhaps the most disruptive element is AI‑Media’s Hardware‑as‑a‑Subscription (HaaS) offering, which converts a capital‑intensive purchase into an operational expense. By spreading costs over time, smaller broadcasters and digital‑first producers can adopt advanced AI encoding without large upfront budgets. AI‑Media plans tiered pricing from $1,200 to $3,500 per month, aligning cost with channel bandwidth and feature set, a model that could reshape hardware budgeting. The company’s collaborations with NVIDIA, Grass Valley and AudioShake further validate the ecosystem, promising tighter integration with GPU‑accelerated inference and next‑gen audio processing. If the subscription model gains traction, it could accelerate industry‑wide migration to AI‑enhanced live workflows.