New instance details view


We added a new handy tooltip which displays info about the underlying compute instance used in Experiments, Jobs, Notebooks, and Deployments. This tooltip includes the complete specs of the instance, including CPU and GPU as well as a link to the details provided by the provider.

Collapsible Gradient Notebooks banner

A popular request for Gradient Notebooks was to free up screen real estate by making the banner at the top of the screen minimizable. You'll now see a small arrow in the top right-hand corner of the screen to collapse the banner. To expand the banner, just click the arrow again. Enjoy!


Gradient now includes the ability to tag entities for improved navigation, search, and management of your ML workflow. Tags can be created and updated in both the user interface and CLI.

You can add tags to several elements including Projects, Experiments, Models, and Deployments.

Gradient YAML Config File

The YAML config file simplifies your work from the command line by storing default arguments for different tasks. If you find yourself repeatedly typing the same commands and want the configuration to be reusable (and with integrated version control), this tool is for you.

Instead of typing out each argument every time you want to run something from the Gradient CLI, you can save your preferred default arguments in the config.yaml file.

Read more on our blog here.

What’s New 🚀 in 7.2.5

UX & Performance

The performance of various has been optimized to load faster & the design of a couple modals couple (upgrades, accounts) have been improved.


There is now an option to receive an email after billing runs each month. Find this under Team->Billing->Recent Invoices.


The native app now supports QUIC by default - the latency experienced
during streaming a virtual desktop is minimized without incurring any
hits to performance


Added social profiles, the ability to make notebooks public/shareable to your friends & a free GPU. These features are in closed beta at the moment, reach out to support to have it enabled.

Continue to make improvements to the CLI (

Bug fixes to model processing where the summary for tensorflow models was not getting populated correctly.