ChatGPT has recently stunned the world with its novel image generation powers, joining the growing list of AI models that can create incredible images from a single text prompt. Like ChatGPT, Gemini can create images with a simple command, and Gemini also lets you perform powerful edits.
Separately, OpenAI, Google, and other companies have models that let you create mind-blowing videos with text prompts. Some of them allow the use of images in the prompt, so you can make AI movies featuring specific settings or characters.
Finally, we’ve seen AI models that can take a single photo and then animate it. Some can even match voice to the resulting animation, including actual recordings from humans or AI-generated audio. These models aren’t publicly available yet.
If you’re aware of all these AI developments in recent years, you won’t be surprised to hear that Gemini will let you create short, 5-second clips starting from one of your photos. You might need access to a brand-new phone to use such features before anyone else.
It’s not a Galaxy phone, where a Google AI feature like Circle to Search was first available. And it’s not a Pixel phone, where you’ll find all the newest Gemini features by default.
Instead, Honor partnered with Google to offer the new image-to-video feature first. It’ll debut with the Honor 400 series, which launches on May 22nd. The AI Image to Video feature was developed using technology from Google Cloud, which is essentially Gemini tech, and Honor will be the first to use it.
The AI Image to Video tool is based on Google’s Veo 2 model. As you can see in the examples above and below, the AI can create videos out of anything, whether it’s one of your photos or a well-known painting.
The feature is built into the Gallery app on the upcoming Honor 400 phones and is very easy to use. You won’t have to enter a text prompt alongside the image you want to turn into a video. The AI handles it automatically, which is great.
Since the AI will create videos randomly, you won’t be able to abuse the feature to create fake clips of people doing who knows what. That helps reduce the risk of generating potentially misleading clips with AI.
The AI Image to Video tool will be free to Honor 400 Series buyers for the first two months. You’ll be limited to 10 video generations per day.
Honor told The Verge that users will eventually need some kind of Google subscription, but didn’t provide specifics.
Presumably, you’ll need to sign up for Gemini Advanced, Google’s cheapest premium Gemini tier. Current subscribers can make videos with Veo 2, but are limited to text-to-video generation. Image-to-video support is coming to Veo 2 in the future. For now, it’s available to “approved users” only and costs $0.50 per second of video.
Months after Google launched Circle to Search on Galaxy S24 and Pixel devices, it expanded the feature to other Android devices. I’d expect the same thing to happen with the image-to-video tool that the Honor 400 Series is introducing.
With Google I/O 2025 just one week away, I’d expect Google to talk about these Gemini capabilities before Honor unveils the Honor 400 Series.
It’s unclear how the image-to-video feature will work once it rolls out to Gemini users. I’d expect it to support both text and image input so you can guide the AI to make the clips you want. I’d also expect some kind of safety measure, like a permanent watermark, to be part of the feature.
The samples above that Honor provided are impressive, but they don’t include watermarks to show they were created with AI. For more examples, check out the following clip.
The post Gemini AI will soon be able to turn any photo into a video appeared first on BGR.
