AI search engine Perplexity joined the busy AI video market and launched a new tool for premium subscribers on August 12th. This feature allows Perplexity Pro and Max users to create 8-second video clips with sounds from a simple text prompt.
The move will intensify fierce competition in the AI video space and place companies against giants like Google, Microsoft and Openai. By adding this tool, Perplexity aims to give paid users more value and compete in one of the most active areas of technology.
Confusion enters AI Video Arena
The launch has been officially confirmed by CEO Aravind Srinivas. “This new feature is currently exclusively for Perplexity Pro and Max subscribers, and mobile users will need to update their app to the latest version.” This feature is based on previous experiments from Perplexity, such as a video generation tool for X users, but it shows the first full integration for subscribers.
Available in the latest mobile app versions, this tool generates 8-second clips with audio. Pro Plan offers standard access, but according to Perplexity, Max subscribers will get “improve quality” and “higher rate limits” and create a clear value proposition for its top tier.
A crowded field of Titans and Vandals
Perplexity's new tools are already full of powerful rivals and are projected to be worth more than $2.5 billion by 2032. Each major player carries out clear strategies in the races that dominate the generated video, creating complex, fragmented landscapes.
We set up an early benchmark by integrating Google's VEO 3 model sync audio and lip sync. “We've emerged from a quiet era of video production.” The tool has been deployed globally to AI Pro subscribers ($19.99 per month) in over 159 countries.
For a large number of creators, Google offers an AI Ultra plan for $249.99 a month, bundled with massive storage and AI credits. This clear two-tier strategy aims to capture both the mainstream creative market and the professional studio segment.
Microsoft went another path, leveraging its partnership with Openai to offer powerful SORA models for free through Bing Video Creator. This freemium approach, which offers users fast set-to-number works, puts pressure on competitors to justify costs.
This field also includes specialized startups. Known for its artistic image generation, Midjourney launched the V1 video tool to bring movement to still images. In contrast, Runway is an ALEPH model designed to edit existing footage, focusing on post-production.
Alibaba's open source WAN 2.2 model is also an important challenger, using a sophisticated mix of architecture. The ability to generate 720p videos on consumer-grade GPUs allows a much wider audience of developers and researchers to access high-quality AI videos.
The battle between business models and ethics
This fierce competition has led to a wide range of business models. Google offers VEO 3 via AI Pro and Ultra subscription plans, but offers a pay-per-developer API for $0.75 per second to developers targeting both consumers and enterprise clients.
This multifaceted strategy is in contrast to freemium access from Microsoft and Prperxity. The market is still experimenting to find the most sustainable approach to these computational expensive tools, balancing user acquisition and profitability.
Philosophical differences also shape the landscape. Elon Musk's Xai intentionally sought controversy by launching the Grok Imagine generator in “spicy mode” that allows for the creation of NSFW content that includes partial nudes.
The move is directly opposed to the strict content filters of rivals, and coincides with the absolutism of mask-free things. It has attracted acute criticism, particularly following the previous controversy with the Groke model. “Free speech belongs to humans, not artificial intelligence.”
Navigating copyright and safety minefields
Beyond features and prices, the entire industry tackles critical legal and ethical challenges. Most importantly, the unresolved copyright issues that came to mind when Disney and Universal filed a groundbreaking lawsuit against the Mid Journey.
The lawsuit accuses AI Labs of training models on protected intellectual property without permission. In a dull statement that captured the heart of the conflict, said Horacio Gutierrez, Disney's legal advisor. “Copyright infringement is copyright infringement, and the fact that it is being done by an AI company does not infringe it.” The case could reconstruct the way in which all AI models train.
In response to growing concerns about deepfakes and false alarms, businesses are implementing safety measures. For example, Google embeds SynthID digital watermarks in all VEO 3 outputs to ensure transparency and helps identify AI-generated media.
However, these solutions are not silver bullets. Independent academic research at the University of Maryland finds that watermarks may be vulnerable to manipulation, highlighting the ongoing technological weapons race between generation and detection.
Ultimately, the AI video space is a chaotic and rapidly evolving field where complex business and ethical issues collide with technological innovation. For creators, this means an explosion of new possibilities. As filmmaker Darren Aronofsky commented on Veo 3, “Now is the moment to explore these new tools and shape them for the future of storytelling.”
