Plutonic Rainbows

I discovered that my app was failing to display the generated video because I was incorrectly extracting the video URL from the Fal.ai API response. Initially, my code assumed the video data was inside a property called data (i.e., final_obj.data), but in reality, the final result was returned directly as a plain dictionary in final_obj with the structure {"video": {"url": "..."}}. Once I logged the final API response, I realised I needed to use final_obj directly to extract the video URL. This change fixed the issue, and now the correct URL is passed to the template, allowing the video to display as intended.

Another image-to-video model, this time wan-i2v which claims to be the next evolution in video generation.

Built upon the mainstream diffusion transformer paradigm, Wan2.1 achieves significant advancements in generative capabilities through a series of innovations, including our novel spatio-temporal variational autoencoder (VAE), scalable pre-training strategies, large-scale data construction, and automated evaluation metrics. These contributions collectively enhance the model’s performance and versatility

Trying to figure out a way to open links in new tabs.

Pretty incredible results with this new model. It responds very well to prompting.

Snow falling on a fairground in Vienna.

Generated result is ()here. The only issue right now is that generation fees have dropped so the queue time is very long.