Finn McKenty

Kling 3 is here, the followup to the popular Kling 2.6 AI video generation model. After spending the last 48 hours running it through the wringer, I’m ready to call it: this is arguably the most capable general-purpose video model available right now. I’d consider it state-of-the-art, overall. On par with Veo 3.1, and possibly better in some ways. But is it right for your workflow, and your budget? Let’s get into the details and find out. My starting image for the tests What’s New in Kling 3?

Need to change the lighting in an image after you’ve shot it? Nano Banana Pro is surprisingly good at it. We’re talking about results that are on par with a high-end retoucher, but accessible on your devices 24/7. The best part? If you’re already using Nano Banana, you don’t need another subscription. You just need to know how to talk to it. But here’s the secret: getting the best results isn’t about some fancy “secret prompt.” It’s about understanding photography basics.

Higgsfield just dropped this new feature that promises to turn your text prompts into motion graphics. It’s built in partnership with Anthropic (the folks behind Claude), and it seems to be leveraging the power of Remotion (although I’m not 100% sure about that), which has already generated a lot of buzz in the dev community. The promise? “Real-time” motion design control. Prompt, and you get polished motion graphics back. The reality? Well, I took it for a spin so you don’t have to guess.