Google AI Studio Now Lets You Add YouTube Video Links

In just a couple of days, Google unveiled its new model, Gemma 3, announced native image generation in Gemini 2.0 Flash, and shipped a few more advancements to its AI products.

However, among the slew of updates, one feature that stood out was the ability to add a YouTube video link and ask Gemini to summarise the video on Google AI Studio. Logan Kilpatrick, senior product manager at Google DeepMind, shared a demo of how that works on X.

Introducing YouTube video 🎥 link support in Google AI Studio and the Gemini API. You can now directly pass in a YouTube video and the model can usage its native video understanding capabilities to use that, with just a link! 🚢 pic.twitter.com/4jeNVmWtgx

“We’re introducing the ability to pass YouTube links through the API in AI Studio and have the model automatically understand and parse them,” Kilpatrick said. 

In the demo video, he pasted the link for Google’s new Gemma 3 AI model in the Google AI Studio, where the video preview was rendered with the token information as 75,401 tokens. Next, he added a prompt, “Summarise this video”.

The documentation for the feature stated, “The Gemini API and AI Studio support YouTube URLs as a file data part. You can include a YouTube URL with a prompt asking the model to summarise, translate, or otherwise interact with the video content.”

There are some limitations to the feature at the moment. One cannot upload more than eight hours of YouTube video per day, and send one video link per request. Also, the videos need to be public, not private and unlisted.

A Reddit user tried using the feature to analyse where a person looked in the video at a particular timestamp, and it was able to correctly identify that. AIM tried the feature, and it seemed to do a good job.

While this sounds like a time-saving feature, will it affect the views for creators in the long run? Only time will tell.