OpenAI Launches Sora 2 Video Generation Model to Take on TikTok, Instagram

In Focus

• OpenAI has launched Sora 2 AI model about a year since unveiling the first Sora model
• Sora 2 is an advanced AI video generation model for creating short-form videos
• OpenAI launched Sora 2 alongside a TikTok-like iOS social media app called Sora
• The social media app is powered by the Sora AI model

OpenAI has launched its most advanced AI video generation model, Sora 2. According to NDTV, OpenAI launched Sora 2 alongside a new iOS social media app called Sora. The TikTok-like app is powered by the Sora 2 AI model.

Creating More Immersive, Realistic Videos

Sora 2 AI video model launches nearly a year after OpenAI released the first Sora model. OpenAI said Sora 2 marks a significant step towards boosting AI’s ability to produce immersive, realistic video content. The latest AI model can generate sound effects, synchronized dialogues, as well as accurate, realistic motion.

The new AI video model will be accessible free of charge via OpenAI’s new social media app. However, one has to get a Sora app invite to access the AI video model. The TikTok-like app allows users to develop high-definition videos with audio through text prompts.

Unique “Upload Yourself” Feature

OpenAI’s Sora 2 video app comes with a unique tool called Cameo that allows users to upload videos of themselves. The AI model uses its high character consistency to place the character in different settings.

Users can post these videos and their followers can like, comment, and share them. While scrolling through videos, users can use the Cameo feature to take the character in a video and generate another video.

“We first started playing with this “upload yourself” feature several months ago on the Sora team, and we all had a blast with it. It kind of felt like a natural evolution of communication from text messages to emojis to voice notes to this,” OpenAI said in a blog post.

A GPT-3.5 Moment for Videos

The biggest improvement in Sora 2 relates to its native audio generation, which is at par with Google’s Veo 3. OpenAI called the new model “the GPT-3.5 moment for video”, adding that Sora 2 is capable of handling complex movements.

The main features of Sora 2 video app are

• Vertical scroll, swipe-based design like TikTok, YouTube, and Instagram Reels
• AI-powered video generation
• Cameo feature that lets users upload themselves
• Short-form video content only
• Algorithm-driven feeds like personalized recommendations

Just recently, Meta debuted Vibes, an AI-based video feed that allows users to watch short videos generated through in-house AI models.

Highlights of OpenAI’s Sora 2 Launch

• Sora 2 can generate videos involving complex movements
• The Sora 2 model is accessible via the Sora iOS app
• The social media app is an invite-only platform available in the U.S. and Canada only

OpenAI launches Sora 2 barely a week after bringing parental controls to ChatGPT. The AI startup said it considers the wellbeing of teens as critical and that its new AI model comes with “default limits” on the number of generations that teenage users can view in their feeds per day.

“We’re also rolling out with stricter permissions on cameos for this group,” the company said in its blog post, adding that besides “automated safety stacks, we are scaling up teams of human moderators to quickly review cases of bullying if they arise.”

OpenAI also said that it was launching with Sora parental controls through ChatGPT, enabling parents to “override infinite scroll limits, turn off algorithm personalization, and manage direct message settings.”

With the latest launch, OpenAI is redefining how content is generated and shared online. This move places it in direct competition with tech giants like Meta, Google, and ByteDance.

OpenAI has added an agentic shopping system called Instant Checkout on ChatGPT, allowing U.S. users to make in-chat purchases from Shopify and Etsy.