Review: Runway AI Gen-4: Blazing Speed and Mind-Blowing Video Generation

We generated three videos with Runway's Gen-4 Model. Click play to watch.

For those unfamiliar (which i doubt anyone is by now), Runway AI has been making waves in the generative AI space, particularly with its impressive suite of tools for multimedia creation. And let me tell you, Gen-4 feels like a significant leap forward.

I've been hands-on with various AI image-to-video tools before, often left impressed by the potential but sometimes frustrated by the execution speed or the lack of fine-grained control. However, my initial foray into Gen-4 has been nothing short of exhilarating. Runway is the type of AI tool that really gives you taste of the future power and potential of AI video generation. We are not far away from entire movies being generated using Runway and then being distributed at a commercial level.Gen-4 feels like a significant leap forward, heralding what Runway themselves are calling "a new generation of consistent and controllable media."

To truly test its capabilities, I decided to generate three distinct video clips, each presenting its own unique set of challenges. First, I envisioned a classic fantasy scene: a wise wizard, staff in hand, walking alongside his majestic white horse through a breathtaking valley. The result was surprisingly evocative, capturing the serene journey with a gentle, almost ethereal quality.

Next, I wanted to push the boundaries a bit, attempting a more nuanced and emotionally complex scene. I prompted Gen-4 to create a shot of an actress in her dressing room, sitting before a mirror with a glass in hand, her expression hinting at a subtle melancholy. While capturing the precise emotional depth I envisioned proved to be a fascinating challenge (as often is the case with AI and nuanced human emotion), the generated clip offered an intriguing glimpse into the potential for creating more introspective and character-driven moments.

Finally, for a more straightforward test, I prompted a close-up shot of a man smiling directly into the camera. This, Gen-4 handled with impressive ease, rendering a warm and engaging moment with convincing facial expressions.

The creative power at my fingertips felt immense across these varied scenarios. Simply typing in a detailed prompt and hitting "generate" felt almost magical. But what truly blew me away was the speed. Gone are the days of waiting what felt like an eternity for a few seconds of grainy footage. Gen-4 rendered these clips with remarkable swiftness. I was genuinely impressed by how quickly these imagined scenes materialized before my eyes. It felt less like waiting for a machine and more like collaborating with a hyper-efficient digital assistant.

The quality of the generated videos, even with the more challenging emotional scene, was also noteworthy. The wizard's journey felt authentic, the actress's contemplative moment held a certain intrigue, and the smiling man was undeniably engaging. It's clear that Runway has made significant strides in the realism and coherence of its video generation.

However, my exploration wasn't without its learning curves. While Gen-4 is incredibly powerful, I quickly realized the importance of precise prompting. Simply throwing in a vague idea won't necessarily yield the exact result you're envisioning. I found myself iterating on my prompts, adding more specific details about the camera movement, the actions within the scene, and the overall mood I was aiming for. But this is to do more with my prompting skills than the software itself. On Runways website and other Youtube videos you can clearly see an array of spectacular video content and Runway's ability to precisely generate consistent characters, locations, and objects across scenes.

For instance, with the wizard scene, initially, the horse's gait wasn't quite right. By refining my prompt to include "smooth, steady walk" and specifying the camera tracking the pair, I achieved a much more satisfying and picturesque result. Similarly, with the actress, I experimented with terms like "soft lighting" and "slightly downcast gaze" to nudge the AI closer to the melancholic mood I was aiming for. This highlights a crucial aspect of working with generative AI: it's a collaborative process that requires a degree of linguistic artistry to guide the AI towards your desired outcome. You need to be the director, and the prompt is your script.

After being thoroughly impressed by Gen-4's creative potential and speed, my attention naturally turned to the practicalities: the pricing. Runway AI offers a tiered subscription model designed to cater to different user needs:

• Basic (Free): This is a great starting point for anyone curious to dip their toes into Runway's capabilities. It offers limited credits (100 with 25 per 5 second video generation) and 720p video exports, allowing you to experiment without any financial commitment.

• Standard ($12/user/month billed annually): Stepping up to the Standard plan unlocks more advanced video generation features, unlimited projects, the ability to train custom AI models, and HD video exports. This seems like a solid option for individual creators or small teams starting to integrate AI into their workflow.

• Pro ($28/user/month billed annually): The Pro plan provides a significant boost in credits, the exciting ability to create custom voices for lip-sync and text-to-speech, and the coveted 4K video exports. This tier likely targets more professional users with demanding project requirements.

• Unlimited ($76/user/month billed annually): For those who foresee heavy usage, the Unlimited plan offers exactly that – unlimited video generation alongside all the features of the Pro plan, including access to their base image generation model, Frames.

• Enterprise: For larger organizations with specific needs and potentially multiple users, Runway offers custom enterprise solutions with tailored pricing.

It's important to note that the Standard, Pro, and Unlimited plans support a limited number of users per workspace (initially 5 or 10, with additional charges for more editors), and the monthly credit allowances are typically shared within that workspace.

Final Thoughts:

My initial experience with Runway AI's Gen-4 has been genuinely exciting. The speed of execution is a game-changer, and the creative possibilities feel vast, even when tackling more nuanced and challenging prompts. While mastering the art of precise prompting is key to unlocking its full potential, the learning curve seems manageable. The tiered pricing structure offers options for various levels of usage, from casual exploration to professional production.

Runway AI Gen-4 feels like a significant step towards a future where AI-powered video creation is more accessible and efficient than ever before. I'm eager to continue exploring its capabilities and see how it evolves, especially in capturing more subtle human emotions. If you're a content creator, filmmaker, or simply someone fascinated by the cutting edge of AI, developments in Runway AI is definitely worth keeping a close eye on.