Beamr Joins VAST Cosmos to Demo Smaller AI-Enhanced Archive Files at IBC

Beamr has joined the VAST Cosmos Community as a technology partner and will demonstrate an AI workflow for video archives at IBC 2026 in Amsterdam. The collaboration combines enhancement and compression to address a basic archive problem: higher-quality video can also create more data to store and deliver.

One pipeline, several archive tasks

The September 11–14 demonstration brings together the VAST AI Operating System for Broadcast, NVIDIA Video Super Resolution and Beamr’s content-adaptive bitrate technology. The end-to-end pipeline runs on NVIDIA RTX PRO GPUs and can be deployed in cloud or on-premises environments.

What the demonstration is designed to do

• Upscale lower-resolution video to 1080p or 4K, or improve quality at the existing resolution.
• Convert older media assets to HEVC or AV1 codecs.
• Apply Beamr’s content-adaptive bitrate technology to the enhanced output.

Up to 50% smallerFile size versus standard encoding

Beamr says its content-adaptive bitrate technology can make the resulting video up to 50% smaller than standard encoding.

The quality-versus-cost test

Compression is the economic part of the pitch. Upscaling footage can increase the number of bits a broadcaster must retain and send; Beamr says its technology can reduce the output size while fitting existing delivery chains. Its up-to-50% figure is a company performance claim, not a result established across every archive.

Beamr also says AI enhancement will not benefit every asset equally. The company recommends evaluating results on a customer’s own material with real viewers before moving a file into delivery, making perceived quality as important as the technical pipeline.

A new partner route to market

VAST introduced its unified Cosmos partner program in February as a framework for technology, cloud and channel partners to build, validate, sell and deliver AI-ready solutions on the VAST AI Operating System. Beamr and VAST also plan joint go-to-market work around complementary video and data workflows.