They really never miss. Higgsfield just added their best image model yet, becoming one of the most powerful AI platforms. Here’s a quick test using the new image model + Keyframes + 3D Rotation effect. Prompts and quick breakdown below:

Yesterday I showcased how you can use the new Higgsfield Soul model combined with camera movements, but I think it works even better with VFX transitions:

Genesis_0 – Awakening | Official Short Film 🩸 This short was created using the most advanced AI video and voice models. But here’s the catch: I limited myself to using only one image model, Flux Kontext Max. Did it work? Watch and find out. Enjoy!

Few more tests with Hailuo 2. It's doing a great job at following long prompts and a very specific order of actions. -> Character moves weapon, followed by a visual effect, ending with the object's destruction Both models do well, although Hailuo is ~10x cheaper

last cig I promise

WIP - Metamorphosis

WIP – Metamorphosis II A few more tests with Flux Kontext and Kling AI 2.1 Enjoy the loop!

Introducing LetzAI Chat: Your creative partner that knows what you love ✨

Visual World building with Flux Kontext and Kling AI Elements. Using the right ref image is key. You can extend from here and pretty much get all angles you want of the character and location. If the consistency is not there yet, you can easily swap characters with Kling AI.