First look: Wan 2.1 from Alibaba vs. Kling AI

🚀 Alibaba gets in on the act: First test of the Wan 2.1 AI video model

I'm starting to lose track - it feels like a new AI tool or function appears every day. But of course I had to take at least a quick look at the next player: Wan 2.1 from Alibaba. Soon to be open source. Sounds exciting, doesn't it?

I kept it simple for my test:
✅ Snapped a few Midjourney pictures
✅ Provided with my VideoGPT + prompt
✅ Run through Kling AI & Wan 2.1

And what came out of it? 🤔
👉🏻 Extremely dynamic videos
👉🏻 Very “exaggerated” implementation of the prompts
👉🏻 Definitely has potential

I assume Wan 2.1 simply needs a different prompting, will give it a second try soon.

🎬 Have any of you already had experience with Wan?

👇🏻Here you can find two example prompts:

*A low, high-speed tracking shot chases a blazing Formula 1 car hurtling through narrow city streets, tires screeching around each sharp turn. The camera shakes slightly to convey the raw power and speed, capturing the driver’s intense focus through the open cockpit. The surrounding buildings whip past in a blur, neon reflections shimmering on the sleek car’s polished surface. As the vehicle drifts around a final corner, sparks fly from the undercarriage. 8K cinematic, ultra-detailed motion blur, dramatic lighting.*

*A dramatic low-angle tracking shot starts from the floor, slowly rising to reveal a young woman standing defiantly, her burning wings igniting with intensity. The flames crackle, flaring outward as embers scatter through the air. The camera slightly circles her, emphasizing the contrast between her delicate attire and the raw power of the fire. She takes a deep breath, her eyes glowing subtly, as the flames pulse in sync with her energy. Shot in 4K HDR, cinematic lighting with high contrast, slow-motion fire effects, and immersive depth of field.*