HeadsUpAI

HeyGen Open Sources Smart Captions and Local Background Removal for HyperFrames

HeyGen, an AI video generation platform, expanded its open-source HyperFrames framework with 17 customizable caption components and local background removal. HyperFrames allows AI agents to compose videos by writing HTML and CSS, which the engine then renders into deterministic MP4 files.
New caption components
17 styles
Background removal
Local, studio-grade
License
Apache 2.0
Core technologies
HTML, CSS, GSAP
System requirements
Node.js 22 and FFmpeg
Total catalog items
78 blocks and components

This update challenges paid AI video editors that treat smart captions and background removal as premium moats. By moving these features into an open-source environment, HeyGen is lowering the floor for high-volume content automation. It builds on the framework's existing catalog of shaders and 3D blocks to provide a free toolkit for video-as-code.

You can install the new caption styles using the npx hyperframes add command. The background removal tool runs on your own hardware, avoiding cloud API costs and keeping assets local. These components are designed for agentic coding workflows, enabling AI assistants to autonomously build and style professional-grade social clips.

View the full update on github.com
HeyGen
HeyGen
@HeyGen
X

“Smart Captions” - the most expensive but basic feature in video apps So we open sourced 17 captions in HyperFrames Free. Fully customizable. One command to install. $ npx hyperframes add captions Plus local background removal — free, studio-grade Make your content POP! https://t.co/oBnladetXm

48retweets266likes
View on X

Still wondering? A few quick answers below.

HyperFrames is an open-source video rendering framework that allows developers and AI agents to create videos using HTML, CSS, and JavaScript. Unlike traditional video editors that use a visual timeline, HyperFrames treats video as code, enabling automated and deterministic rendering where the same input always produces the identical MP4 output.

Smart Captions are a collection of 17 customizable components that provide high-quality text overlays for video compositions. Users can install these components, such as the per-word clip-path wipe reveal, using a single terminal command. Once added, AI coding agents can autonomously build these captions into video files and customize their appearance through code.

Yes, the entire HyperFrames framework and its new caption components are fully open source under the Apache 2.0 license. This allows individuals and companies to use the tool commercially at any scale without paying per-render fees, seat caps, or meeting company-size thresholds that are common in proprietary video rendering software.

While both tools drive headless Chrome for deterministic video rendering, HyperFrames uses standard HTML and CSS for authoring instead of React components. HyperFrames does not require a build step, allowing compositions to play as-is in a browser. Additionally, HyperFrames is fully open source, whereas Remotion requires a paid license for many commercial use cases.

HyperFrames now includes a studio-grade background removal feature that runs locally on the user's hardware. This allows creators to strip backgrounds from video assets without sending data to a cloud service or incurring external API costs. The local execution ensures faster processing and better privacy for professional video production workflows.

Share this update