The last few years have brought a series of changes to the industry. New tools, reduced budgets, and daunting client expectations in the wave of AI. Lived experience and professional standards now stack against a very different technological ecosystem.

Line 21 has made a decision with that in mind. We are making our platform free for human captioners when the workflow is human text in, human text out. If a human captioner or notetaker needs to send their output live captions to a browser, overlay, or supported API token destination then they can do so without cost.

For us, this is a practical decision that respects the foundational role human captioners have in the accessibility industry. Accessibility tools should work for the professional, with simple stepping stones to translation and video services, rather than absorb humans as a byproduct of bygone days.

As workflows change, and requests for increasing translation flexibility and integration from clients grow, the conversation around AI becomes more heated. In many ways it has never been a more popular time to be accessible; but with that rising wave of "I want captions at my event" comes the sometimes painful learning curve that the highest stakes still need human hands.

What free access includes

For human captioners, Line 21 will make human text delivery free across supported outputs where the workflow remains fully human-led.

This includes:

• Captions to browser
• Captions to overlay
• Captions to supported API token destinations such as Zoom, Microsoft Teams, and YouTube (if a destination is not available, and you need it to be — talk to us)

The boundary is simple. A human captioner provides the text. Line 21 delivers the text.

Where captioners need additional services, such as machine translation or streaming, Line 21 has them available.

The free platform access is focused on the core human captioning workflow.

Building with captioners, not around them

We also know that simply opening access is not enough. We understand that one captioner's keystroke definitions are not another's — and we work with global teams, who work in very unique ways between regions, languages and team sizes.

We — the human builders — expect to learn from you. We expect some quality of life interface changes to be needed. We expect connector requests (Advantage Eclipse, we heartily salute your 10 day integration time). We expect regional adjustments.

We expect some workflows to be familiar, and as captioners become comfortable with our wizards in the workshop to ask the "what if" question of design to suit them.

The question people are already thinking about

There is one part of this announcement that deserves direct attention. AI.

Line 21 uses AI across the platform. We deploy 12 engines across 6 services. We provide services such as ASR, machine translation, and other automated workflows for clients who need them.

For human captioners, that fact can create understandable concern. We're so open with how we use it; our co-founders enjoy linguistic and technical challenges so much they once provided machine translations in Klingon.

Unlike others out there, even more heinously, we expose our engines. At every point you interact with AI in Line 21's platform, you know what you're dealing with. You know we have tested it, and when we deploy something new, we share our excitement with you.

Because of our rampant transparency on these points, we are happy to confirm we do not train AI on any input from our users. You cannot trust our system if we are not clear on data, training, consent or control. We know very well of the reality of this fear, observing the Zoom training controversy with raised brows ourselves.

The industry has already seen the manifestation of anxiety in a very real way, courtesy of that Zoom incident. Many captioners had to look again at where their words were going, what platforms could do with meeting content, and whether human-generated captions might become training material for systems that would later compete with the people who produced them. For professionals whose work depends on trust, confidentiality, skill, and consent, that went to the heart of control.

Line 21 is not a video conferencing platform. For human captioners using the free human-in, human-out workflow, Line 21 is a text delivery platform. The captioner sends human-generated text. Line 21 outputs that text to the chosen destination. We are not hosting the meeting, processing the video, listening to the room, or using a captioner's work to train a model.

No AI engine in Line 21 can be turned on by accident, or learns from human captioners who use the platform.

Any contact with AI inside Line 21 has to be chosen with intent. It cannot happen as a hidden background process. It cannot happen because someone clicked the wrong button in a human captioning workflow.

If a captioner is using Line 21 for human text in, human text out, that is the workflow.

We also turn training data off with the AI engines and vendors we work with for speech recognition. In some cases, providers offer significant discounts to encourage training data to be enabled. We do not accept those discounts. The principle matters more than the saving.

Control needs to be designed into the platform, and that is our approach.

For more detail on vendor retention, AI training, and third-party processors, see Data handling & retention.

Why are we doing this?

The captioning industry — along with all the others — is being asked to absorb a lot at once.

Clients are learning new terminology, along with expectations and service comparisons that are not equivalent. Human providers are being asked to explain the difference between accessible and approachable.

That can become tiring. It can also make people wary of platform announcements that sound generous at first reading.

We understand the concern. When a technology company says something is free, many people immediately wonder where the cost has been moved. Is the user now the product? Is the free tool a short-term hook? Will the useful features be placed behind a paywall later? Will the workflow become difficult to leave once people depend on it?

For this human captioning workflow, the answer is simple. Line 21 is making it free and unlimited because we can, and because it is a useful thing to do.

There is no hidden training exchange. There is no requirement to use ASR. There is no requirement to use machine translation. There is no requirement to send us video.

Line 21 benefits from increased usage of the platform. That is true, and we are not trying to disguise it. By following the rule of universal design, everyone benefits. We create a wider base of professionals who may choose to use paid services such as machine translation or streaming when those services are genuinely useful to them.

The future of accessibility should not be built by pushing skilled people out of the process. It should be built by giving them better tools, clearer control, and fewer unnecessary costs.

An invitation to human captioners

For human captioners who are curious about trying a new delivery option, or who simply want to reduce the cost of the tools they already need, this is an open invitation. Use it for a job. Test it with a colleague. Try it with a browser output, an overlay, or a supported platform connection. If it helps lower your operating costs, that is enough reason for us to offer it.

Line 21 works with human providers globally, and we expect this free access to grow through use. Some improvements may be small. Some may come from a captioner trying the platform on an ordinary workday and really wanting a big red button.

How to get free access

Human captioners who want to use the free caption flow can register for a Line 21 organisation at line-21.com.

Once registered, email hello@line-21.com and let us know that you would like the free human captioning flow enabled. We will turn it on for human text in, human text out traditional workflows.