Runway Unveils Advanced AI Video Generator

# gen-4 # runway

AI startup Runway just dropped what they're calling one of the most high-fidelity AI-powered video generators out there. They've named it Gen-4, and it's now available to both their individual and enterprise clients. Runway boasts that Gen-4 can keep characters, locations, and objects looking the same across different scenes, hold together coherent world environments, and even recreate elements from new angles and positions within those scenes. In their blog post, Runway explains that Gen-4 can take visual references and instructions to whip up new images and videos that stick to consistent styles, subjects, and settings, all without needing any extra training or fine-tuning. > Gen-4 sets a new standard for video generation and is a marked improvement over Gen-3 Alpha. It excels in its ability to generate highly dynamic videos with realistic motion as well as subject, object and style consistency with superior prompt adherence and best-in-class world… pic.twitter.com/w9ACO5boJ7 > > — Runway (@runwayml) March 31, 2025 Backed by big names like Salesforce, Google, and Nvidia, Runway offers a range of AI video tools, including Gen-4. They're up against tough competition from the likes of OpenAI and Google, but they're carving out their niche. They've even signed a deal with a major Hollywood studio and are pouring millions into funding films that use AI-generated video. Runway says Gen-4 lets users create consistent characters across different lighting conditions by using a reference image. To set up a scene, you can upload images of subjects and describe how you want the shot to look. > Using visual references, combined with instructions, Gen-4 allows you to create new images and videos with consistent styles, subjects, locations and more. Allowing for continuity and control within your stories. > > To test the model’s narrative capabilities, we have put together… pic.twitter.com/IYz2BaeW2U > > — Runway (@runwayml) March 31, 2025 According to their blog post, "Gen-4 excels in its ability to generate highly dynamic videos with realistic motion as well as subject, object, and style consistency with superior prompt adherence and best-in-class world understanding." They also claim that "Runway Gen-4 \[also\] represents a significant milestone in the ability of visual generative models to simulate real-world physics." Like other video-generating models, Gen-4 was trained on a huge number of video examples to pick up on patterns and create synthetic footage. Runway's keeping tight-lipped about where the training data came from, worried about losing their edge. But this secrecy could also land them in hot water with IP-related lawsuits. For instance, Runway's currently facing a lawsuit from artists who claim that the company and other generative AI outfits trained their models on copyrighted artwork without permission. Runway's defense? They're leaning on the fair use doctrine. It's still up in the air whether this will hold up in court. The stakes are pretty high for Runway, especially since they're reportedly looking to raise funds that would value them at $4 billion. According to The Information, they're aiming for $300 million in annual revenue this year, thanks in part to new products like an API for their video-generating models. No matter how the lawsuit turns out, generative AI video tools are shaking things up in the film and TV world. A 2024 study by the Animation Guild found that 75% of film production companies using AI have cut, merged, or eliminated jobs. The study predicts that by 2026, over 100,000 entertainment jobs in the U.S. could be disrupted by this technology.

Related article

Comments (46)