Google Cloud today announced an update to its Cloud Video Intelligence API that can provide users a transcript of conversations that took place in a video, one of a series of AI-related updates.

In addition to transcripts for videos, a text detection tool was also added to the Cloud Video API, along with object tracking for more than 500 common objects such as dog, flower, or car.

Pricing for second-generation tensor processing unit pods was also shared today and can be seen on the Google Cloud AI & ML pricing website.

Tensor processing units are custom-made chips for the fast training or inference of AI models. Second-generation TPU chips for cloud computing made their debut last year and became generally available in July.

Also out today from Google is Kubeflow Pipelines, a machine learning workflow meant to make ML easier for developers and data scientists.

A Google AI Hub also launched today in alpha. The AI Hub, designed to be a one-stop shop for getting started with AI, includes things like popular datasets on Kaggle and TensorFlow embeddings.

In time, Google wants the AI Hub to also become a  place to find trending AI models from the wider ecosystem.

“We eventually want AI Hub to be a place where third parties can also share information and turn it more into a marketplace over time,” Google Cloud senior director of product management Rajen Sheth told VentureBeat in an interview at Google offices in San Francisco.