Join our daily and weekly newsletters to receive the latest updates and exclusive content on our industry-leading AI coverage. Find out more
Lightricksthe Israeli company behind the viral photo editing app Facetuneis launching an ambitious effort to shake up the generative AI landscape. The company today announced the release of LTX Video (LTXV), an open-source artificial intelligence model that can generate five seconds of high-quality video in just four seconds. By making its video model freely available, Lightricks takes direct aim at the growing dominance of proprietary AI systems from tech giants like OpenAI, Adobe AND Google.
“We believe that core models will become a commodity and you can’t build a real business around core models,” said Zeev Farbman, co-founder and CEO of Lightricks, in an exclusive interview with VentureBeat. “If startups want to have a serious chance of competing, the technology must be open, and you want to make sure that people from the best universities around the world will have access to your model and add capabilities to it.”
With real-time processing, scalability for long video formats, and a compact architecture that runs efficiently even on consumer-grade hardware, LTXV is able to bring professional-grade generative video technology to a wider audience – an approach that has the potential to disrupt the industry status quo.
How Lightricks uses open source software to challenge AI giants
Lightricks’ decision to release LTXV as open source it’s a calculated risk designed to differentiate the company in an increasingly crowded generative AI market. ZIts two billion parameters, model is designed to run efficiently on commonly available GPUs such as Nvidia RTX4090while maintaining high visual fidelity and motion consistency.
The move comes at a time when many leading artificial intelligence models — starting with OpenAI DALL-E to Google Picture—are locked behind APIs, requiring developers to pay for access. Lightricks, on the other hand, assumes that openness will foster innovation and adoption.
Farbman compared LTXV’s release to Meta’s open source release Llama language modelswhich quickly gained popularity in the AI community and helped Meta establish itself in the OpenAI-dominated space ChatGPT. “The business case is that if the community embraces it, if people in academia embrace it, we as a company will benefit tremendously,” Farbman said.
Unlike Meta, which controls the infrastructure on which its models run, Lightricks focuses solely on the model itself, working with similar platforms Face Hugging to make it available. “We do not intend to make money on this model at this time,” Farbman emphasized. “Some people are going to deploy it locally on their hardware, such as a gaming PC. It all depends on adoption.”
Lightning-fast AI video: Breaking speed records on consumer hardware
What makes LTXV stand out is its speed. The model can generate five seconds of video – 121 frames at 768×512 resolution – in just four seconds on an Nvidia processor H100 GPUs. Even on consumer-grade hardware like RTX4090LTXV delivers near real-time performance, making it one of the fastest models of its type.
This speed is achieved without any loss of quality. Models Transmission transformation architecture ensures velvety motion and structural consistency between frames, eliminating key limitations of previous video generation models. For smaller studios, independent creators and researchers, the ability to iterate quickly and generate high-quality results on inexpensive hardware is a game changer.
“When you wait several minutes for a result, it’s a terrible user experience,” Farbman said. “But when you get feedback quickly, you can experiment and iterate faster. You create a mental model of what the system can do, and this unlocks creativity.
Lightricks also designed LTXV to support longer video productions, offering creators greater flexibility and control. This scalability, combined with tiny processing times, opens up up-to-date opportunities for industries from gaming to e-commerce.
In gaming, for example, LTXV can be used to enhance the graphics in older games, transforming them into visually stunning experiences. In e-commerce, the speed and efficiency of the model can enable companies to create thousands of ad variations for targeted A/B testing.
“Imagine casting an actor – real or virtual – and tweaking the visuals in real time to find the best performance for a specific audience,” Farbman said.
From a photography app to the power of artificial intelligence: Lightricks’ bold market play
With LTXV, Lightricks is positioning itself as a disruptor in an industry increasingly dominated by a handful of tech giants. It’s a bold move for a company that started as a mobile app maker and is best known for Facetune, a consumer photo editing app that became a global hit.
Lightricks has since expanded its offerings by acquiring a Chicago-based influencer marketing platform Popular pays and launch Studio LTXan AI-powered storytelling platform for professional creators. LTXV’s integration with LTX Studio is expected to expand the platform’s capabilities, enabling users to generate longer, more energetic videos with greater speed and precision.
However, Lightricks faces significant challenges. Competing with heavyweights such as: Adobe AND Autodeskthat have deeper pockets and an established user base won’t be effortless. For example, Adobe has already integrated generative artificial intelligence into its solutions Creative Cloud package, which gives it a natural advantage among professional users.
Farbman recognizes the risks, but believes open source innovation is the only real path forward for smaller players. “If you, as a startup, want to have a chance to take on the giants, you need to ensure that the technology is open and embraced by academia and the broader community,” he said.
Why open source can win the AI video generation race
The release of LTXV also highlights the growing tension in the AI industry between open source and proprietary approaches. While closed models offer companies tighter control and monetization opportunities, they risk alienating developers and researchers who don’t have access to cutting-edge tools.
“Currently, we are observing, among other things, that diffusion models are becoming an alternative paradigm to the classic methods of operation in computer graphics,” Farbman explained. “But if you actually want to build alternatives, APIs are definitely not enough. You have to give people – academia, industry, enthusiasts – models that they can tinker with and create amazing up-to-date ideas.
Lightricks plans to release LTXV on both GitHub AND Face Huggingwith an initial “community preview” phase for testing and feedback. The model will eventually be released under the name OpenRAIL Licenseensuring that derivatives remain open for academic and commercial apply.
The stakes are high for Lightricks. The company is betting not only on the success of LTXV, but also on the broader adoption of open AI models in a field increasingly dominated by closed ecosystems.
“The future is bright for open models,” Farbman said confidently.
Time will tell whether this vision will be realized. But by making its most advanced technology freely available, Lightricks is sending a clear message: in the race to define the future of video AI, openness and collaboration may prove to be the greatest competitive advantage.