Corporate training video has been bad for thirty years for reasons that have nothing to do with technology. Someone writes a script that reads like a policy document, a presenter delivers it flatly to camera, and nobody watches past ninety seconds. Synthesia Academy's genuine contribution is that it spends most of its time on that problem rather than on the AI. The Academy is free, requires only a free account, and is built around short courses with video lessons produced using Synthesia's own avatars.
The centrepiece is FOCA, a scripting framework covering how to open, how to structure, and how to write for an audience rather than for an approver. It is not complicated and it does not pretend to be, but it gives people who have never scripted anything a repeatable structure, and the difference between a video written to that structure and one written without it is enormous. I would happily recommend that course to somebody making videos with a camera and no AI involved at all, which is not something I expected to write about vendor training. The wider material covers planning, storytelling, visuals, audio and audience first scripting, with exercises that require you to build and publish actual videos rather than answer questions about them.
That insistence on producing something is the right call. Video skills do not develop through watching, and a course that ends with three published videos beats one that ends with a quiz. Some courses offer completion certificates, and there is a Synthesia Certified track for people who want a more formal credential. Using the product to teach the product is also an honest move.
You are watching AI avatars explain how to make videos with AI avatars, which means you form an accurate impression of the output quality within the first two minutes. Some people will find it slightly uncanny and decide the tool is not for them, and that is a legitimate outcome that the Academy does not try to hide. I prefer that to a slickly produced human led course selling a product that looks nothing like the course. The limitations follow from what this is.
It is narrow. Everything assumes avatar and synthetic voiceover video, so you will learn nothing about lighting, camera work, directing a real person, or the many situations where a synthetic presenter is the wrong choice. It is also unambiguously a funnel. The training is free because Synthesia is not, and once the course ends every skill you have acquired requires a subscription to use.
That is a reasonable trade and it is obvious enough that nobody should feel misled, but it is worth stating plainly. Synthesia Certified deserves the same treatment. It is a product credential. Inside a company that has standardised on Synthesia it identifies the person who knows what they are doing, which has some value.
Outside that context it means very little, and it should not be described on a CV as though it were a video production qualification. The omission that bothers me most is disclosure. AI avatar video raises real questions that the training walks past. Should an audience be told the presenter is synthetic.
What consent does a company need before using an employee's likeness as an avatar. How do customers actually react when they find out. These are live issues, they vary by jurisdiction, and organisations are getting them wrong right now. Training aimed at people about to deploy this across a company ought to cover them, and it would cost Synthesia nothing to do so beyond a slightly less frictionless funnel.
My 3.8 is for free, well made, practically useful training with a scripting framework that genuinely improves the thing most people are bad at, held back by narrow scope, a credential worth little outside its own ecosystem, and a conspicuous silence on disclosure. Take the FOCA course regardless. Take the rest once you have decided you are buying.