Cardboard Turns Raw Footage Into Finished Video With One Prompt

The Macro: Video Editing Is the Last Manual Bottleneck in Content

Every marketing team I have talked to in the last year has the same complaint. They can write copy fast. They can generate images instantly. They can ship landing pages in hours. But video? Video still takes days. Sometimes weeks. The editing process involves agencies, review cycles, revision rounds, and a level of manual labor that feels almost quaint compared to how quickly everything else moves.

The video editing software market is dominated by heavyweight tools. Premiere Pro, DaVinci Resolve, Final Cut Pro. These are incredible products built for professional editors. They are also wildly overpowered for the marketing manager who just needs three 30-second clips from a webinar recording. The gap between “full professional editing suite” and “I just need a clean cut with captions” is enormous, and most teams fill it with either expensive agencies or underpowered tools like Canva’s video editor.

AI video tools have been arriving steadily. Descript made text-based editing mainstream. Opus Clip automated short-form clipping. Runway and Pika went hard on generative video. But the agentic editing space, where you give AI your footage plus a goal and it produces an edit, is still wide open. Multimodal LLMs can finally understand video content well enough to make editorial decisions, and that changes the game for tools like Cardboard.

The Micro: Give It Footage, Give It a Goal, Get a Cut

Cardboard positions itself as an agentic video editor for growth teams and serious creators. The workflow is: upload your raw footage, describe what you want (“3 variants,” “30-second hook,” “testimonials compilation”), and Cardboard produces a first cut in minutes. Not a rough assembly. A clean, editable first pass that you can iterate on.

The product runs in the browser, which means no downloads and no rendering on your local machine. Under the hood, it uses Claude Sonnet 4.6 for the multimodal reasoning layer, which handles understanding what is happening in the footage and making editorial decisions about pacing, cuts, and structure.

Feature-wise, Cardboard handles silence removal, color grading, caption generation, voiceover integration, and smart trimming. It has specialized modes for talking heads, vlogs, montages, podcast clips, launch videos, and explainers. Live collaboration is supported, so teams can review and iterate together.

The thing that separates this from Opus Clip or Descript is the agentic layer. You are not making editing decisions and having AI execute them. You are describing an outcome and letting the agent figure out the editorial approach. That is a fundamentally different interaction model, and it is only possible because multimodal models have gotten good enough to reason over video semantically.

Saksham Aggarwal and Ishan Sharma are the founders. They came through Y Combinator’s W26 batch. Pricing starts at $60 per month with tiered plans available.

The competitive question is whether Descript adds agentic features before Cardboard builds enough market share. Descript already has the user base, the brand, and the integrations. Cardboard’s advantage is starting agent-first rather than retrofitting agency onto an existing editor.

The Verdict

Cardboard is attacking the right pain point for the right audience. Marketing teams do not want to learn DaVinci Resolve. They want to ship video content at the same speed they ship everything else. An agentic editor that takes raw footage and produces a polished first cut in minutes is exactly what that audience needs.

At 30 days: how good is the first cut, really? If teams still spend two hours cleaning up what the agent produces, the time savings evaporate. The quality of the initial output is everything.

At 60 days: does the agent learn from your feedback? If I keep adjusting pacing and the agent never picks up on my preferences, it becomes a one-trick tool rather than a real workflow partner.

At 90 days: can it handle long-form content? A 30-second clip from a webinar is straightforward. A 10-minute product demo edit from an hour of raw footage is a completely different challenge.

The $60/month price point is aggressive and smart. It undercuts agencies by orders of magnitude and sits comfortably below what teams spend on Premiere Pro seats that nobody fully uses. I think Cardboard has real potential if the output quality holds up at scale.