> For the complete documentation index, see [llms.txt](https://blog.gooey.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://blog.gooey.ai/gemini-live-3.1-gpt-image-2-and-bulk-eval-upgrades.md).

# Gemini Live 3.1, GPT Image 2 & Bulk Eval Upgrades

Product Updates April 2026

## Stop Runs Anytime

No more waiting around for a run to finish if you change your mind. We've added a Stop button to cancel in-progress runs, and background tasks are revoked immediately so nothing keeps chugging away in the background.

<img src="https://3205006002-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FleYcqBx5FRZcVr3wI4f4%2Fuploads%2FVNqi6ycPw3pxQxnFBcff%2Funknown.png?alt=media&amp;token=eb1b2b9e-6b99-451d-aa1d-f30d71627003" alt="" height="353" width="624">

## Clearer Credit Errors

Running low on credits? You'll now get a dynamic error message that tells you exactly what happens when a run fails due to insufficient credits, no more guessing.

## Gemini Live 3.1 in Voice Deployments

Your voice agents just got smarter. We've added support for Gemini Live 3.1 in LiveKit voice deployments, bringing the latest model capabilities straight to your calls.

## Deploy Any Workflow as an LLM Tool

You can now deploy any workflow as an LLM Tool, complete with a WhatsApp country code selector. Turn your workflows into building blocks other agents can call.

## GPT Image 2 Support

We've added support for GPT Image 2, giving you another top-tier option for image generation workflows.

<figure><img src="https://3205006002-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FleYcqBx5FRZcVr3wI4f4%2Fuploads%2FpJOpU9YfEh2oOpCWqiLC%2Fimage.png?alt=media&amp;token=b1f5967c-42c1-48b3-b148-65c48688c735" alt=""><figcaption></figcaption></figure>

Try it here

{% embed url="<https://gooey.ai/compare-ai-image-generators/?run_id=97deft2semvr&uid=fm165fOmucZlpa5YHupPBdcvDR02>" %}

## Smarter Bulk Eval

Bulk Eval just got a serious upgrade: a code editor for prompts, graph captions, and a new "lower is better" grading option for metrics where smaller numbers mean better results.

<img src="https://3205006002-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FleYcqBx5FRZcVr3wI4f4%2Fuploads%2FkN7X4fiQBsjvbgp7GPuO%2Funknown.png?alt=media&amp;token=cf6ce870-e183-4f2b-8651-021e996e4999" alt="" height="315" width="624">

<img src="https://3205006002-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FleYcqBx5FRZcVr3wI4f4%2Fuploads%2FDzOd5lsBbXofdi0bJFpL%2Funknown.png?alt=media&amp;token=d2c12f8a-1acc-44a9-ba31-0a8e0988cc2a" alt="" height="581" width="554">

## Dynamic LLM Tool Search

LLM tools can now be discovered dynamically, making it easier to find and connect the right tool for your agent on the fly.

## Last Login, at a Glance

Managing a team? The workspace members view now shows a "Last login" column, so you can see who's active and who isn't in one glance.