Demos
See it working.
Real footage, real people, dated. No renders, no AI signing. If it is here, it ran. Then: how your screens get it, with one API.
In plain words
These are videos of the real thing working. Some are older. Under each one we say what it shows and what is still missing from it — like human captions or a signed version. Further down: how a platform or city connects to it.
This page in sign language
International Sign for the global site; your national sign language on your country page. Signed by a Deaf creator, credited by name.
Sign language
A signer appears in sync with the content, in the viewer's own sign language.
Also shows: captions, audio description.
The access layer on a real screen: sign language, captions and audio description delivered over content the viewer already has.
Audio description
Description that carries the story, for blind and low-vision viewers.
A short clip with an audio description track carrying the story.
What you will not find here
If it is here, it ran.
- No concept renders. Every frame is a real screen, a real room, a real person.
- No AI-generated signing. Every signer is a human. Creators are credited by name as their videos ship.
- No roadmap videos. Things that do not exist yet are not demonstrated. They are not here.
- Honest labels. Each video says whether its captions are human yet, whether a transcript exists, and whether a signed version exists. Before launch, every one gets all three.
Demonstrate, don't announce.
Want to see it on your own content? Talk to us — or read on for how the integration works.
For platforms, broadcasters, cities and services
You have captions. Deaf users still can't follow.
Fixing it is one API call, not a content operation. The community makes the tracks. You deliver them — over content your viewers already have.
140–230 words a minute
Captions run at 140–230 words per minute; median Deaf comprehension is around 100–140 — and comprehension falls as caption rate rises.
Receipt: Jensema 1998; Burnham 2008; Tyler 2009
The median deaf school-leaver in the best-resourced deaf-education system reads at about a fourth-grade level.
Receipt: Traxler 2000
What the API returns
Three kinds of track. One request.
In plain words
Your app asks for a piece of content. We answer with what exists for it: sign language, captions, audio description. Your player shows it on top of your video. We never touch your files.
Sign language tracks
In the viewer's own sign language, with more than one interpreter where the community has made them. There is no universal sign language; the API knows this.
Captions
Human captions, in the written language of the market. Timed to the content.
Audio description
Description that carries the story, for blind and low-vision viewers. Delivered the same way.
Our tracks render over content your viewers already have.
The overlay principle: we never re-host, copy or modify your content. Your player, your rights, your files — plus the access layer the community made for it. One platform, three strands: content made by the community, code that delivers it, and a community that is credited and paid.
Where it goes
Every screen your users already use.
- Streaming and on-demand video
- E-learning
- Government portals and livestreams
- Banking and service apps
- Smart TV
- News
Who starts here
Four ways in.
- Governments and cities Accessible services, livestreams and public information. Start on the policy page.
- Broadcasters and platforms Sign language and description on the content you already publish.
- Telecoms Accessible-service bundles and relay obligations, served through one integration.
- Enterprises Customer service and training that reach Deaf and blind customers and staff.
The marketplace
A rope has two ends.
A rope has two ends — that is the marketplace.
Creators supply. Your users are served. The community is the supply chain — born warm, not cold-started.
Marketplaces die on the cold start; ours was born warm.