Runway has introduced Gen-4.5, its latest video generation model, targeting creators, filmmakers, enterprises, and anyone seeking advanced video AI tools. The company is gradually rolling out Gen-4.5 to all users, with public availability set in the coming days across all subscription plans. The model supports a global user base via Runway’s platform, offering both individual and organisational access.
Introducing our new frontier video model, Runway Gen-4.5. Previously known as Whisper Thunder (aka) David.
— Runway (@runwayml) December 1, 2025
Gen-4.5 is state-of-the-art and sets a new standard for video generation motion quality, prompt adherence and visual fidelity.
Learn more below. pic.twitter.com/2J945S4bQt
Gen-4.5 delivers high visual fidelity, physical accuracy, and creative control, setting a new standard in AI-generated video. It achieves 1,247 Elo points on the Artificial Analysis Text to Video leaderboard, outperforming all competing models. Technical advancements include improved pre-training data efficiency, new post-training strategies, and optimised inference running on NVIDIA Hopper and Blackwell GPUs. The model excels in generating cinematic and photorealistic outputs, handling complex scenes and nuanced emotions, with consistency in object motion and scene detail. All prior control modes, such as Image to Video and Keyframes, are carried over. However, common video model challenges remain, such as occasional lapses in causal reasoning or object permanence.
Founded as a pioneer in AI-powered video, Runway previously launched Gen-1, one of the first public video generation models, and has since led in model advancements and usability. The company’s ongoing collaboration with NVIDIA drives further optimization and ensures that Gen-4.5 delivers breakthrough quality and creative power for a broad range of professional and creative users.
