Midjourney remains our go-to for AI-generated imagery. It's fast, stylized, and great for concept art, thumbnails, and visual direction. We often use it to create creative, brand-consistent imagery in place of stock photos.

Midjourney now supports basic animation, allowing you to animate an existing image—either created within Midjourney or imported externally. The process is simple and fast, ideal for prototyping. Midjourney's main advantage is that it’s built for creative professionals: the output is well composed and visually striking, and the system is fast with generous usage limits.

Google's Veo 3 can animate a still image, or generate a complete video directly from a text prompt—something Midjourney doesn’t support. Veo 3 excels at realism, with a stronger grasp of physics and fewer strange visual artifacts. Veo 3 is also the only tool (at least for now) that can generate synchronized audio, including dialogue that matches up with the video.

In our workflow, we treat Midjourney and Veo as separate tools for different needs: Midjourney when we want stylized visuals or concept-driven animation, Veo when we want realism, motion accuracy, or sound.

Both tools require manual effort and taste. Midjourney still has an AI-generated look that can be hard to override, and Veo outputs are limited by slower generation times and a lack of creative variety.

Neither tool is a replacement for professional video work. But they are good enough to generate early-stage assets, and even finished results depending on the project's requirements and expectations.

One limitation of both tools is resolution. Midjourney animations output at 480p, and Veo currently delivers 720p by default. These aren’t high enough for full-width web banners or fullscreen playback.

We use Topaz Video AI 7 to upscale output when higher resolution is needed. It supports frame interpolation, motion-aware upscaling, and artifact reduction—giving us clean, sharp results at 1080p or 4K. This step makes AI video assets more usable in production environments without quality concerns.

The current generation of AI video tools is far from complete, but already useful. While they’re not yet a match for traditional video production, they’ve earned a place in our creative workflow, often serving as a more flexible and brand-consistent alternative to generic stock footage.