Remember two years ago? When Gen-1 launched and we all pretended that morphing, blurry mess was "cinema"? Good times. Well, Runway has decided we've suffered enough. They've just released Gen-4.5—a new frontier for video generation with state-of-the-art motion quality, prompt adherence, and visual fidelity.
The Numbers Don't Lie
Gen-4.5 dominates the Artificial Analysis Text-to-Video leaderboard under the mysterious codename "Whisper Thunder" (aka David). With 1,247 Elo points, it's currently looking down at everyone else. Yes, that includes Veo 3.1. Yes, that includes Kling 2.5. It's surpassing all other models in the space.
Control Freak? This is for You
Runway knows that creators are control freaks (it's okay, this is a safe space). Gen-4.5 isn't just about typing a prompt and praying to the algorithm gods.
They're bringing over all the toys: Image to Video, Keyframes, Video to Video. It gives you the power to handle a wide range of aesthetics while maintaining a coherent visual language.
Translation: You might actually be able to make a short film where the main character doesn't change faces three times in five seconds.
So, if you’re ready to see if the hype matches reality, head over to https://app.runwayml.com/