Pika Labs is about to challenge Google Veo 2 with upcoming version 2.1 video model

· 1 min read
Pika

Pika Labs recently announced the upcoming release of their new video generation model, version 2.1. On their official account, they shared a sample video along with an early access link. Early feedback from partner creators suggests that this model could potentially compete with Google’s Veo 2 in terms of quality. Currently, this is regarded as a top-tier video generation model, though it remains exclusive to trusted testers.

0:00
/0:05

"Cyberpunk hacker robot" benchmark via Pika 2.1

At TestingCatalog, I had the opportunity to test this model as well. I experimented with both text-to-video and image-to-video features, and the results were impressive. The text-to-video capability stands out as one of the best I’ve tested for my benchmark prompts, which you can find here. The image-to-video feature is equally remarkable, capable of animating images with exceptional quality. It managed to handle tasks that other models struggled with, such as animating cartoon characters.

The exact release date of this model is still unclear, but it’s expected to have a significant impact on video generation. Once the model is widely released, creators will be able to publish videos generated during early access, and we can look forward to seeing some truly impressive creations.