Introducing
Runway Gen-4.5:
A new frontier for video generation.

State-of-the-art motion quality, prompt adherence and visual fidelity.

December 1, 2025

by Runway

Introduction

Runway Gen-4.5 is the world's top-rated video model, offering unprecedented visual fidelity and creative control. It produces cinematic and highly realistic outputs while providing limitless creative freedom and precise control over every aspect of generation.

Gen-4.5 Announcement

Already in use by:

Introducing Runway Gen-4.5

Two years ago, we released Gen-1, the first publicly available video generation model. It enabled an entirely new form of creative expression and a new product category. Since then, we've led the industry in making video models more powerful and controllable, from significant base model improvements to new controls and general in-context capabilities.

Runway Gen-4.5 pushes the frontier of video generation even further. It represents significant advances in both pre-training data efficiency and post-training techniques. Gen-4.5 sets new standards for dynamic, controllable action generation, temporal consistency and precise controllability across diverse generation modes. With 1,247 Elo points, Gen-4.5 currently holds the top position in the Artificial Analysis Text to Video benchmark, surpassing all other models.

Gen-4.5 maintains the speed and efficiency of Gen-4, delivering breakthrough quality without compromising performance. Available at comparable pricing across all subscription plans, Gen-4.5 makes world-leading video generation accessible to creators and organizations at every scale. We’ll also be bringing all existing control modes (Image to Video, Keyframes, Video to Video and more) to Gen-4.5.

Evolution of Video Generation

Gen-4.5 achieves unprecedented physical accuracy and visual precision. Objects move with realistic weight, momentum and force. Liquids flow with proper dynamics. Surface details render at great fidelity. And fine details like hair strands and material weave remain coherent across motion and time.

Complex Scenes

Intricate, multi-element scenes rendered with precision.

Prompt:

Camera is panning and trucking across a kitchen table with different objects placed on the counter. The light is low, it comes through a window, casting shadows and dappled light.

Prompt:

A continuous shot with no scene cut. We see water coming out of a garden faucet filling up a rusty metal bucket. Right below the bucket is a perfect folded white paper boat with a small yellow flower. The boat rides down a small river of water and enters a house, ending the journey in front of a broken TV.

Prompt:

As the camera is panning to the right, we see a pool with a man inside, a rolled up rug, a mirror, a cloud, a TV, a cooked chicken, among other various random objects.

0 of 0

Detailed Compositions

Precise placement and fluid motion for both objects and characters.

Prompt:

A polar bear sits huddled inside a clear frozen ice block, strapped securely to a small silver utility trailer. The camera follows the trailer pulled by a cord at eye level, then pans to the right to reveal a goat pulling the trailer.

Prompt:

The video opens with a realistic, naturally lit scene. The overall color palette is muted with cool undertones. The video shows a large whale flying through a winter mountain range. The weather is mostly cloudy with small bits of blue sky peeking through.

Prompt:

Orbiting a snowman in New York. Suddenly, a man bursts out from inside of its center and walks away. Handheld documentary style, natural camera shake, indie film aesthetic.

0 of 0

Physical Accuracy

Realistic physics with believable collisions and natural movement.

Prompt:

Anthropomorphic cactus person hugs red balloon person, and the balloon person pops.

Prompt:

A scuba diver in the bottom of the ocean firing a machine gun. The camera is handheld and imperfect, sometimes off center.

Prompt:

An open watermelon, a chrome colander, a long skinny cactus in a green ceramic pot and a parrot balancing on top of each other, wobbling and shaking.

0 of 0

Expressive Characters

Nuanced emotions, natural gestures and lifelike facial detail.

Prompt:

A cinematic scene highlights a young woman with striking features, captured in a close-up that emphasizes her intense gaze. The camera focuses on her face, showcasing her platinum blonde hair.

Prompt:

High-intensity, visceral, horror-style metamorphosis in a single continuous shot. Close-up side profile of a regular person in warm, dim indoor lighting, camera positioned tightly on their face.

Prompt:

A brunette woman with intense and serious facial expression approaches the camera, then raises her arm revealing a gun. She then fires the gun several times, pointed toward the camera.

0 of 0

Gen-4.5 can handle a wide range of aesthetics, from photorealistic and cinematic to stylized animation, while maintaining a coherent visual language.

Photorealistic

Visuals indistinguishable from real-world footage with lifelike detail and accuracy.

Prompt:

A young woman with straight blonde hair and a freckled complexion sits quietly, looking up through her lashes with a vulnerable expression.

Prompt:

Continuous underwater shot with no scene cuts. Camera is traveling forward underwater, with different objects floating in the water. The light is low, it comes through the surface of the water.

Prompt:

A trucking wide side profile shot following a flying squirrel rising through the air with its arms and webbed wings outstretched.

0 of 0

Non-photorealistic

Stylized, expressive motion with artistic freedom unconstrained by realism.

Prompt:

A dapper country fox wearing a straw boat hat and a suit is riding his motorbike on a dirt road through the countryside. A hyper-detailed, stop-motion-inspired cinematic style.

Prompt:

A dense, verdant jungle world made up of small block-like pieces. We see a chameleon running through the 3D world, the camera in and out of focus.

Prompt:

Behind view of a lonely wanderer walking through a dark and mysterious forest. Classic fantasy film featuring richly saturated early color photography paired with theatric styling.

0 of 0

Slice of Life

Everyday scenes and environments with authentic, true-to-life detail.

Prompt:

The scene shows a the TV that has a 90's dog cartoon. A hand comes in to turn it off. The camera pans downwards to reveal "Runway" written in cereal in a bowl with milk, then back upwards toward the fridge to reveal "Whisper Thunder" written in alphabet magnets.

Prompt:

A helicopter chases a speeding race car down a dirt road. The sky is delicately colored with light blue with hints of orange, suggesting it might be sunrise or sunset.

Prompt:

The video opens with a realistic, naturally lit scene showing two open hands, palms facing upwards. There is a blue stone in the left hand and a red one in the right. Another hand enters the frame and takes the blue stone away.

0 of 0

Cinematic

Emotionally powerful visuals with striking depth and cinematic polish.

Prompt:

Close-up shot. A left hand grips the edge of a mirror. We can see an old man in the mirror as he tilts it up and down.

Prompt:

A kid on the back of a Mandrill, next to a group of Mandrills running on a plain. Cinematic, fast motion blur, handheld documentary film style, low angle.

Prompt:

In a moonlit forest blanketed with swirling fog, a single uninterrupted shot trails a man as he cautiously makes his way between gnarled trees. The beam of his flashlight cuts through the fog.

0 of 0

High-performance

Built on NVIDIA

Gen-4.5 was developed entirely on NVIDIA GPUs across initial R&D, pre-training, post-training and inference. We collaborated extensively with NVIDIA to push the boundaries of what's possible in video diffusion model optimization, from training efficiency to inference speed.

Inference runs on NVIDIA Hopper and Blackwell series GPUs, delivering optimized performance without compromising quality.

“

This is an incredibly exciting time for video and world models.

We’re proud that Runway built their groundbreaking video and world model on NVIDIA GPUs, and are thrilled to see Runway revolutionize the video generation industry.

Together, we are partnering to advance the entire lifecycle of AI from pretraining, to post-training and inference.”

Jensen Huang

President and CEO of NVIDIA

Select Early Access Enterprise Partners

Retail & Ecommerce

Marketing & Advertising

Gaming

Despite the leap in capabilities, the model exhibits several limitations common to video generation models:

Causal reasoning: effects sometimes precede causes (e.g., a door opening before the handle is pressed).
Object permanence: objects may disappear or appear unexpectedly across frames (e.g., a cup vanishing after being occluded).
Success bias: actions disproportionately succeed (e.g., a poorly aimed kick still scoring a goal).

These limitations are particularly important in our work on world models, which need to accurately represent the outcomes of actions taken in the environment. We are actively researching ways to address them.

Generated with Runway Gen-4.5