Self-hosted alternatives to hyperframes-media

H

hyperframes-media

Generate speech, transcribe audio

hyperframes-media is an agent skill from HeyGen's HyperFrames project for creating the audio and media assets a composition needs. It covers voiceover from text to speech, background music and sound effects, transcription with timestamps, captions and background removal for transparent overlays. The skill is organized around one audio engine that takes a neutral audio request file and writes audio metadata plus the generated assets, so the different workflows share one source rather than each carrying its own copy. The media commands are tts, transcribe and remove-background. Each one downloads and caches its own model on the first run, and the skill says no API keys are required for those commands. Voice generation can also use a HeyGen, ElevenLabs or other provider, depending on the credentials present, and the skill explains how the provider is chosen. Once the assets exist, the composition consumes and animates the timing data in its HTML, and the skill points to the core HyperFrames skill for placing them. It suits video makers who want a narrated, captioned or background-cleaned clip assembled locally. Because the first run downloads models, expect a delay the first time. On skills.sh, Gen Agent Trust Hub shows a warning, and Socket and Snyk pass. It has about 141,700 installs. Install it with npx skills add https://github.com/heygen-com/hyperframes --skill hyperframes-media.