Synthesia Launches AI-Powered Video Assistant: Transforms Text into Professional Videos

Synthesia, headquartered in London, is a startup that enables businesses to create professional AI videos. Recently, the company announced the launch of its AI Video Assistant, a tool that can convert text-based materials into complete synthesized videos in minutes.

Starting today, paying customers can access this AI Video Assistant. It works on any document or web link, making it easier for teams to create internal and external video content.

While it facilitates content delivery for businesses, it has also raised some concerns. More and more people are paying attention to AI-generated videos and deepfakes.

What can Synthesia's AI Video Assistant do?

Founded in 2017, Synthesia is composed of AI researchers and entrepreneurs from UCL, Stanford University, TUM, and Cambridge University. It offers an end-to-end platform that allows users to create custom AI voices and realistic AI avatars, or choose existing avatars, and use them to create studio-quality videos.

The product has been widely adopted in the corporate environment, but the process of creating these videos may require some effort, including manually writing scripts (or using AI to write scripts) and designing video layouts, selecting avatars, languages, and elements such as logos.

With the new AI Video Assistant, Synthesia simplifies all these steps. Users only need to provide the source of the materials, whether it's a website, text file, Word document, PDF, or creative content, and then select a template that includes the objectives of the materials, the number of scenes, and the language and tone of the spokesperson appearing in the video.

The AI Video Assistant combines the user's raw materials with their objectives, audience, and other features using generative AI. Large language models accept these inputs and generate scripts and related scene layouts from the selected template.

Sundar Solai, the product manager at Synthesia, who led the development of the assistant, said, "This output is generated within minutes and can be transformed into a video. If the script is not satisfactory, users can stop, go back, and regenerate the settings for better output."

Goal of improving content delivery experience

According to Insivia, when people receive information through videos, they can remember 95% of the information, while they can only remember 10% when reading text.

However, even today, many organizations still rely on lengthy and dense manuals and PDF files to deliver valuable content to employees and customers. With the new AI Video Assistant, Synthesia hopes to bridge this gap and enable any individual in any organization (as long as they have appropriate access) to transform complex textual information into studio-quality videos.

Sundar Solai stated that they have been testing this product for several weeks and have noticed significant interest from customers. However, he did not disclose how many companies participated in the early testing or how they used this assistant.

"In addition to enabling users to create videos faster than before, this feature also accelerates Synthesia's ability to make video creation more accessible, including for individuals with limited or no video editing experience," he pointed out.

While this can greatly facilitate content delivery for businesses, it is worth noting that the tool still has limitations in terms of how much text it can convert into videos. Whether it's a text file or a link, Synthesia stated that the assistant currently has a source limit of 4,500 words.

Despite the concerns surrounding deepfake technology, Synthesia continues to focus on significant growth in enterprise users.

The company, which has 300 employees, raised $90 million in June 2023 and attracted over 55,000 businesses, including half of the Fortune 100. One of their clients is the video communication service Zoom, which has increased the speed of creating videos for sales and training purposes by 90%.

"This means that Zoom's subject matter experts no longer need to record their own videos, freeing up 15-20 hours per month that they can allocate to other tasks as part of their work. Zoom saves $1,000-1,500 per month in costs that would have been spent on producing training videos," said Alexandru, Synthesia's Head of Affairs and Policy.

Voica also confirmed that the company is working on speeding up the platform and its next-generation virtual character technology. This will allow them to display more expressions, better sound, and more customization options in a more natural way. This will bring the AI videos on the platform closer to real videos. Other companies competing with Synthesia in this field include Deepbrain AI, Rephrase, and HeyGen.