Google Antigravity gets a Gemini 3.1 Pro upgrade! Initial impressions.

On February 19th, a minor update, Google Antigravity 1.18.3, was released.
The overall description of the changes is quite modest:

New settings screens for models and terminal integration, support for uploading artifacts, and various stability and UI fixes across different platforms.

Here are the key details of the update:

• Gemini 3.1 Pro is now available.
• A "Models" screen has been added to the settings, providing a more visual representation of quota usage.
• A setting has been added to enable or disable terminal integration.
• Support for uploading artifacts from the chat UI.
• Keyboard navigation with up/down arrows for input field history.
• Improved UI responsiveness during chat interactions, including creating dialogs, sending messages, and undoing changes.

Introducing the upgraded Gemini 3.1 Pro brain.

My first impressions of using it:

• It's faster. It's difficult to objectively evaluate, but it clearly "thinks" faster;
• The "tone of the conversation" has changed; in my opinion, it has become "more human";
• The generated artifacts, in Russian, have changed their tone. It's better, using very understandable phrases, it's actually pleasant to read :)

Changes in Antigravity itself:

• "Skills" are now working correctly; they weren't launching for me before;
• Information about used quotas and update deadlines has appeared. Currently, it only shows how much time is left until the reset;
• The interface has become more responsive.

Conclusions

As Google itself writes, the changes in 3.1 are mainly focused on multimodality. It is not yet clear how this will affect coding.
They state that the available context for coding has increased, but it is difficult for me to verify this, as my projects are quite simple.