Synthesia

Synthesia describes itself as the number one AI video platform for business ^[1]. That ranking is the company's own claim rather than an outside measurement, but it says plainly what the product is for: making video for workplaces — training, internal communication, product walkthroughs — without a camera, a studio or a film crew. Instead of filming a person, the user writes a script and an AI-generated presenter, called an avatar, speaks it on screen.

What the platform includes

The product set covers AI avatars, an AI video generator, AI video dubbing, AI video translation, an AI screen recorder and AI motion graphics ^[1]. An avatar can be created from a single photo or a short video ^[1], which is the step that makes the rest possible: once a presenter exists as a model rather than as a recording, any script can be put into their mouth. The generator turns a written script into a finished video with that presenter; the screen recorder captures a computer screen for software demonstrations; motion graphics adds animated text and diagram elements on top.

Dubbing and translation are related but not the same. Dubbing replaces the spoken audio; translation renders the script in another language. The company says its translation covers 140 or more languages while keeping names, terms and pronunciation consistent ^[1]. That detail matters in practice. A product name, a job title or a technical term that drifts slightly from one language version to the next is one of the clearest signs that a video was machine-translated, and keeping such terms fixed is part of what makes a localized video usable inside a company.

An API is offered alongside the video products ^[1], so video creation can be triggered from other software rather than only through the web interface — useful when a company wants to produce a video as part of an existing workflow rather than as a separate task.

Sessions: practice rather than watching

Sessions is a separate product line for interactive roleplays and conversational surveys ^[1]. The roleplay use cases the company lists are sales roleplay, discovery calls, cold calls and manager training ^[1]. In each, a learner talks with an AI character that plays a customer, a colleague or a direct report, and gets to rehearse a conversation that would otherwise have to be practiced on real people. Survey sessions are pitched as conversational surveys at scale for employee insights ^[1], meaning the same conversational format is used to collect answers from staff instead of presenting them with a form.

Who it is for

The industries served include manufacturing, tech and software, healthcare, financial services, professional services and retail ^[1]. The site also claims the platform is trusted by over 90% of Fortune 100 companies ^[1]. The pattern behind that list is organizations with many employees, many languages and a steady need to explain procedures, policies and products.

Where it sits

Corporate video has traditionally been expensive per minute and expensive to change: fixing one sentence could mean assembling people, equipment and a location again. An avatar-based workflow separates the presenter from the recording, so a script edit becomes a text edit. The same property makes multilingual versions cheap to produce, which is why translation sits at the centre of the product rather than at the edge.

The limits follow from the same design. An avatar video is only as good as its script, since there is no performance to carry weak writing, and it depends on viewers accepting a synthetic presenter as a normal part of workplace communication. Sessions points at a different direction: not video to be watched, but conversation to be practiced.