AI Tools
Image AIBeginner

Cardboard

Cardboard, developed by the Cardboard team, a member of the Y Combinator W26 batch, was released on March 24, 2026. It is the first agentic video editor that operates within a web browser. This tool is an AI-based video auto-editing solution that allows users to upload videos and specify their final production goals in natural language text. The built-in AI director then comprehensively and deeply analyzes the video's visual elements and audio flow to independently establish an editing plan and construct a timeline. It utilizes modern browser standard technologies such as WebCodecs and W.

Cardboard, developed by the Cardboard team, a member of the Y Combinator W26 batch, was launched on March 24, 2026. It is the first agentic video editor that operates within a web browser. This tool is an AI-based video auto-editing solution that allows users to upload videos and specify their final production goals in natural language text. The built-in AI director comprehensively and deeply analyzes the video's visual elements and audio flow to autonomously establish an editing plan and construct a timeline. By leveraging modern browser standard technologies such as WebCodecs and WebGL2 pipelines, it organically combines cloud and browser environments to deliver fast and smooth real-time frame rendering speeds without the need for local installation of heavy applications.

Traditional video editing solutions have involved cumbersome and tedious manual work, requiring users to manually cut frames and drag and drop them onto the timeline, which has been a significant barrier to entry for both beginners and busy professionals. Cardboard aims for a more natural interaction, moving away from this mechanical workflow, allowing users to interact as if they were conveying a script and intentions to a professional video director and receiving the finished product. While existing simple smart cut tools only mechanically cut out silent sections based on volume values, this tool understands visual cues from speakers and the semantic context of speech in a multifaceted way, making intelligent and creative editing decisions that do not disrupt the natural continuity of the story. This is a key differentiator.

In particular, this intelligent video summarization and refinement capability can be very usefully applied in the fields of biotechnology research and academia. In long confocal time-lapse videos used for research presentations or as supplementary materials for papers, it can intelligently select and quickly edit and highlight only the important sequences where meaningful changes, such as cell division, are captured. Furthermore, in long academic presentation sessions or educational webinars on experimental protocols, it can uniformly refine frequent silences or audio noise from the presenter and quickly generate deployable video results with high-precision captioning, allowing for the addition of multilingual subtitles with a single click. The initially assembled timeline results can be exported in standard formats such as XML or EDL, ensuring seamless compatibility with Adobe Premiere Pro or DaVinci Resolve in the final mastering stage, allowing it to function smoothly as part of a professional workflow.

💻 System Requirements

🧠RAM

브라우저 하드웨어 가속 가능 용량 (WebGL2 및 하드웨어 비디오 디코딩 가속 호환 GPU 필수)

💾Storage

로컬 캐시 임시 파일 저장 공간 약 500MB~1GB

Installation

4-1. Quick Start

별도의 패키지 설치는 필요하지 않으며, 브라우저를 통해 서비스로 즉시 진입합니다.

open https://usecardboard.com

4-2. 상세 설치

로컬 개발 환경용 라이브러리가 아니므로, Chrome 혹은 WebGL2 가속이 활성화된 브라우저에서 공식 사이트에 접속하여 가입 후 사용합니다. Premiere Pro 및 DaVinci Resolve 등 외부 편집기와 연동 시 Export 탭에서 XML 또는 EDL 메타데이터 파일을 다운로드하여 임포트합니다.

🧬 Bio Use Cases

🔬

🔬 Automated Summarization and Cut Editing of Real-time Microscopy Imaging Timelapse Videos

Upload a 1080p 60fps timelapse video (120 minutes) of cell observation recorded on Cardboard, specify the prompt 'Edit only the cell apoptosis segments at 10x speed', and automatically extract summarized clips within 15 seconds. This process automatically creates a key presentation video for conferences in under 1 minute from the original 2-hour video, reducing conference preparation time by over 90%.

🧬

🧬 Audio Refinement and Automated Caption Generation Pipeline for Biotechnology Academic Webinars

Upload a 2-hour webinar recording, set the silence_threshold parameter of the audio analysis engine to -35dB to remove silent and simple noise segments within 3 seconds, and generate Korean/English automatic captioning scripts with 98% accuracy to quickly publish summary videos for research promotion.

💊

🧪 Protocol Guide Creation through a Hybrid Workflow with Professional Editing Tools

Build a preliminary rough cut of the experimental protocol footage (4K, 24fps) using AI Director, then export it in XML/EDL format and import it into Adobe Premiere Pro and DaVinci Resolve to apply fine color correction, completing the production of a publicly available experimental guide video within 1 hour.

FAQ

What is Cardboard?

Cardboard, developed by the Cardboard team, a member of the Y Combinator W26 batch, was launched on March 24, 2026. It is the first agentic video editor that operates within a web browser. This tool is an AI-based video auto-editing solution that allows users to upload videos and specify their final production goals in natural language text. The built-in AI director comprehensively and deeply analyzes the video's visual elements and audio flow to autonomously establish an editing plan and construct a timeline. By leveraging modern browser standard technologies such as WebCodecs and WebGL2 pipelines, it organically combines cloud and browser environments to deliver fast and smooth real-time frame rendering speeds without the need for local installation of heavy applications. Traditional video editing solutions have involved cumbersome and tedious manual work, requiring users to manually cut frames and drag and drop them onto the timeline, which has been a significant barrier to entry for both beginners and busy professionals. Cardboard aims for a more natural interaction, moving away from this mechanical workflow, allowing users to interact as if they were conveying a script and intentions to a professional video director and receiving the finished product. While existing simple smart cut tools only mechanically cut out silent sections based on volume values, this tool understands visual cues from speakers and the semantic context of speech in a multifaceted way, making intelligent and creative editing decisions that do not disrupt the natural continuity of the story. This is a key differentiator. In particular, this intelligent video summarization and refinement capability can be very usefully applied in the fields of biotechnology research and academia. In long confocal time-lapse videos used for research presentations or as supplementary materials for papers, it can intelligently select and quickly edit and highlight only the important sequences where meaningful changes, such as cell division, are captured. Furthermore, in long academic presentation sessions or educational webinars on experimental protocols, it can uniformly refine frequent silences or audio noise from the presenter and quickly generate deployable video results with high-precision captioning, allowing for the addition of multilingual subtitles with a single click. The initially assembled timeline results can be exported in standard formats such as XML or EDL, ensuring seamless compatibility with Adobe Premiere Pro or DaVinci Resolve in the final mastering stage, allowing it to function smoothly as part of a professional workflow.

When should I use Cardboard?

Cardboard, developed by the Cardboard team, a member of the Y Combinator W26 batch, was released on March 24, 2026. It is the first agentic video editor that operates within a web browser. This tool is an AI-based video auto-editing solution that allows users to upload videos and specify their final production goals in natural language text. The built-in AI director then comprehensively and deeply analyzes the video's visual elements and audio flow to independently establish an editing plan and construct a timeline. It utilizes modern browser standard technologies such as WebCodecs and W.

What is a biomedical use case for Cardboard?

🔬 Automated Summarization and Cut Editing of Real-time Microscopy Imaging Timelapse Videos: Upload a 1080p 60fps timelapse video (120 minutes) of cell observation recorded on Cardboard, specify the prompt 'Edit only the cell apoptosis segments at 10x speed', and automatically extract summarized clips within 15 seconds. This process automatically creates a key presentation video for conferences in under 1 minute from the original 2-hour video, reducing conference preparation time by over 90%.

📄 Official Docs

📝 Update Notes

No update notes yet.

🧪 Related Code of Life

No related Code of Life posts yet.