About CaptionHub

CaptionHub is a software platform for managing the end to end creation of accessibility and translation assets for timed-media (video and audio) – giving its customers perfectly timed and accurate captions, subtitles, voiceover, and sign language for any video or audio asset.

Orchestration and control over the workflow, scale and detail of all people, technology and models involved that ecosystem are managed in CaptionHub for the purpose of accessibility and localisation for any part of the modern enterprise, studio or language vendor - or for anyone creating video for a global audience for that matter.

It uses a combination of best-in-class web technologies including translation systems, video platforms, and AI/LLM models – as well as our own proprietary technologies – to decrease turnaround times and costs.

Custom dictionaries, term-bases and style guides give those models your product names, speakers and house style before a job starts. NVIDIA, Ford, GitHub, Supercell, TED and several of the world's largest broadcasters and studios use it.

Natural Captions, translation memory and real-time QC turn pre-contextualised AI output into captions, subtitles, voiceover and sign language. Finished work goes out as any major timed-text format, burnt-in subtitles, voiceover audio or a live stream, frame-accurate and with quality scores you can report on.

CaptionHub is ISO 27001 and SOC 2 Type II certified, and customer content is never used to train anyone else's model.