Runway just dropped another AI tool right after their video expansion release – a new image generation model called Frames. I tested it, and the outputs are genuinely indistinguishable from real photos.
Frames gives you precise control over the visual style of your images while maintaining photorealism. This means you can generate images that perfectly match specific aesthetics – from 35mm film photos to retro anime – while keeping the quality incredibly high.
What makes this particularly interesting is how it fits with Runway’s other recent releases. Just days ago, they launched major updates to their video tools, including Gen-3’s image-to-video feature. The rapid pace of these releases shows they’re aggressively pushing their AI image and video tools forward.
The key thing about Frames is its consistency. When you pick a style, it sticks to it perfectly across multiple generations. This is crucial for creating sets of images that look like they belong together – something that’s been a pain point with other AI image tools.
For video creators and filmmakers, this tight integration between Frames and Runway’s video tools opens up new possibilities. You can generate still frames in exactly the style you want, then use them as starting points for video generation.
I’ll be testing Frames once it makes it to their API to see how it performs across different use cases. If you’re interested in AI image generation tools, I recommend checking out my previous coverage of other recent developments in this space:
– Flux.1’s surgical precision approach to AI image editing: https://adam.holter.com/flux-1-tools-ai-image-editing-gets-surgical-precision/
– Krea.ai’s comprehensive image and video generation toolkit: https://adam.holter.com/krea-ai-comprehensive-image-and-video-generation-tools/
I’ll update this post with my detailed findings once I’ve had more hands-on time with Frames.
