Yesterday's Top Launches: 1 Tools from January 24, 2026
Google has released a preview of its new Veo32 AI video model and made the current Veo 3.1 version available for public testing.
Yesterday was a quiet one on the launch front, but it brought a significant update from one of the giants that anyone working with new developer tools should be aware of. While we're tracking several promising early-stage projects, Google's announcement is the clear standout from January 24th.
Veo32 Video
The wait for the next generation of AI video took a concrete step forward yesterday with Google's unveiling of Veo32. This isn't a general release yet; it's the official preview of Veo 3.2, their upcoming flagship text-to-video model. The immediate takeaway is that while we wait for the full capabilities of 3.2 to become available, Google has opened up access to the current iteration, Veo 3.1, for everyone to try.
So, what's the big deal? Veo32 aims to push AI-generated video from the realm of short, often surreal clips into something approaching genuine cinematic quality. The promise is longer, more coherent sequences with a dramatic improvement in visual fidelity, lighting, and detail. If you've ever experimented with earlier models and been frustrated by the character inconsistencies or the "dreamlike" wobble that sets in after a few seconds, Veo32 is positioned as the answer. The underlying technology focuses on better understanding the physics and narrative intent behind a text prompt, translating a director's vision into a moving image with fewer artifacts.
Who stands to benefit the most from this? It's a broad spectrum. Independent filmmakers and content creators are the obvious first group. The ability to generate high-quality storyboards, concept scenes, or even complete short films from text descriptions could drastically lower production barriers and costs. Marketing teams could use it to quickly prototype ad concepts without organizing a full shoot. For educators and explainer video producers, it offers a way to visualize complex concepts on demand. It's less of a coding tool and more of a creative and pre-production powerhouse.
An honest observation here is that the hype cycle for AI video is intense, and Google is competing in a crowded field with other well-funded players. The proof will be in the output. While the technical descriptions are impressive, the community will be the ultimate judge of whether Veo 3.2 delivers on its promise of true cinematic control or if it simply produces better-looking short clips. The fact that they are allowing hands-on experience with Veo 3.1 now is a smart move—it builds anticipation and gives users a baseline to compare the eventual leap to 3.2.
Since this is a preview announcement, concrete pricing for Veo32 itself hasn't been detailed. The current Veo 3.1 is accessible through Google's AI Test Kitchen, which typically operates on a credit-based system. We can expect Veo 3.2 to follow a paid, usage-based model similar to other high-end AI generation services once it fully launches. It's a web-based platform built with Next.js, ensuring a smooth, responsive experience without the need for local hardware powerful enough to run such a demanding model.
This launch is less about a tool you can integrate into your pipeline today and more about a clear signal of where this technology is heading. For developers building applications that might incorporate generative video or for creators planning future projects, keeping a close eye on Veo32's rollout is essential.
Quick Links