D-ID’s newest app uses AI to make videos from photographs | TechCrunch

3 min read


D-ID, the Tel Aviv-based startup best known as the tech behind those viral videos of animated family photos, is bringing its AI video technology to a new mobile app, launching today. Originally available as a web platform, D-ID’s Creative Reality Studio allows users to upload a still image and script and then turn that into an AI-generated video. The technology can be used to create digital representations of themselves, historical figures, fictional characters, presenters, or brand ambassadors.

Early use cases the company had been targeting involved corporate training and education, internal and external communication from companies, and product marketing and sales, TechCrunch previously reported.

Now available on mobile, users will download the D-ID app from the App Store or Google Play and then create an account or log in, if already registered. On the selection screen, you can either pick a premade “digital person” that D-ID provides or upload an image from your phone’s photo library. You’ll then enter the text you want the digital person to say, choosing from 119 languages, as well as pick between male and female voice options. You can also choose the tone of the speech — like cheerful, excited, friendly, hopeful, newscast, sad, shouting terrified, unfriendly, whispering, and others.

Image Credits: D-ID

To work, D-ID is leveraging a mix of proprietary and open-source AI technologies to power its platforms, says D-ID CEO Gil Perry.

“At its core lies a foundational model capable of generating video frames based on audio input. All its products are powered by its robust API with the ability to render video at an industry-leading 100 FPS, four times faster than real-time rendering,” he says.

“The core audience is AI enthusiasts, but it is relevant for anyone who wants to post videos featuring digital people, either versions of themselves or of anyone they can imagine – historical characters, fictional characters, ancestors, etc.,” Perry says. “The app enables video generation on the go, reducing the cost of production at scale, and allows the flexibility to produce videos anywhere.”

Image Credits: D-ID

The videos can be up to 10 minutes in length and are subject to the same moderation policies as the web version — meaning, the company is working to ensure the tech is not used for deepfakes that propagate misinformation and disinformation campaigns. The company has also pledged to respect copyright, adhere to AI regulations, and to not knowingly license its technology to what it calls “contentious areas,” including political parties, pornography publishers, terrorist organizations, and guns or arms manufacturers.

Ahead of the mobile launch, the desktop version of Creative Reality Studio has seen 60,000 new sign-ups per day, with over 150 million videos already made using the platform, Perry tells us. In addition, the API for the service has more than 200,000 users.

With its arrival on mobile, D-ID is able to reach a broader market, including users who are mobile-first or mobile-only. This includes both existing users who are interested in using a leaner version of the studio for mobile devices as well as new users from the general public, including consumers.


The new mobile service will be subscription-based, with plans starting at $5.99 per month. Users already subscribed to the desktop version can retain their existing plans. A 14-day free trial is also available.

Though headquartered in Tel Aviv, D-ID has employees in other markets around the work, allowing it to continue operations during the Israel-Hamas war.

“Our teams in Israel and around the world are working to ensure the highest level of support and service,” says Perry. “It should be noted that our team in Ukraine is operating with extra diligence out of high solidarity with the situation in Israel,” he adds.

Image Credits: D-ID

D-ID raised a $25 million Series B last year, following the launch of its Deep Nostalgia project and its spin-off, LiveStory. Those two projects involved animating historical photos of family members, with the latter adding audio. The videos went viral as people used the tech to create talking photos of long-lost loved ones or ancestors. Since then, D-ID has also been working to create a photorealistic interface for ChatGPT as well as its Creative Reality Studio software, as part of its broader AI development efforts.



Source link