Skip to content
IFAI

AI models

TwelveLabs drops Pegasus 1.5 and an AI video co-pilot

Published Reach 47

TwelveLabs just unveiled Pegasus 1.5 and Rodeo at NAB Show 2026, alongside an integration with Autodesk Flow Capture. Pegasus 1.5 extracts time-based structured metadata in a single API call, outperforming Gemini 2.5 Pro by 30% on video segmentation benchmarks. Meanwhile, Rodeo acts as an AI co-pilot that lets editors find, cut, and assemble footage using plain language.

TwelveLabs drops Pegasus 1.5 and an AI video co-pilot
Image: martechseries.com

Ask any video editor about their worst nightmare, and they will probably describe digging through dozens of hours of raw dailies just to find a three-second cutaway.

At NAB Show 2026, video intelligence company TwelveLabs tackled this exact bottleneck. The team announced Pegasus 1.5, rolled out an AI creative assistant called Rodeo, and revealed a native integration with Autodesk Flow Capture.

Smarter video cuts without manual indexing

The marquee update is Pegasus 1.5, which introduces what the company calls Time-Based Metadata Extraction.

Instead of making production teams tag footage by hand or repeatedly re-index massive files, the model lets users define their own schema. With a single API call, Pegasus 1.5 detects scene boundaries and spits out structured metadata paired with exact timestamps.

In early testing, TwelveLabs claims Pegasus 1.5 beat Gemini 2.5 Pro by 30% on aggregate segmentation quality benchmarks. The model is already deployed in live production with a major broadcast network.

Agentic editing and enterprise pipelines

Beyond raw infrastructure, TwelveLabs also stepped directly into consumer software with Rodeo, its first dedicated application for creators.

Rodeo works like an AI co-pilot. Creators can search, slice, and assemble clips purely using natural language prompts, relying on background AI agents to suggest edits and structure timelines without touching complicated scripts.

On the enterprise side, the company teamed up with Autodesk to bring Smart Search and Smart Actions to Flow Capture, the cloud review platform widely used on both Hollywood sets and indie productions. Media assets are now automatically tagged and routed the moment they land in the cloud.

Why it matters

We think TwelveLabs is playing this smart by shifting away from generic chatbot demos and embedding directly where professional editors already spend their days.

For creators and post-production houses, finding specific moments in unscripted footage has long been an expensive, time-draining manual chore. Bringing structured time-based metadata into everyday tools like Flow Capture means editors can spend less time scrubbing timelines and more time on actual storytelling.

From an industry perspective, hyperscalers like Google are pushing multimodal context windows to the limit, but TwelveLabs is carving out defensibility by focusing on industrial-grade precision and downstream workflows. The big question mark is reliability: professional editing leaves zero margin for hallucinated frames or drifted timestamps, so Pegasus 1.5 will need to consistently prove its edge on live, messy film sets.

Sources (1)

  1. [1]
    martechseries.com(opens in a new tab)Originalmartechseries.com

Similar events

Past happenings like this one, for context.

Best of AI, daily