Skip to main content
This update brings Gemini-powered image generation, image editing capabilities, and the ability to cancel long-running text generations.

Gemini Image Generation (Nano Banana)

Added Gemini 2.5 Flash Image (codenamed Nano Banana) for text-to-image generation. You can now create images using Google’s Gemini models directly in your workflows.

Image Editing

The image generation capabilities now include image editing support. Edit and modify existing images as part of your AI workflows.

Cancelable Text Generation

You can now cancel text generation while it’s in progress. This is especially useful for:
  • Stopping long-running generations
  • Testing and iterating faster
  • Managing resource usage

Long-Running Task Support

Redesigned the text generation pipeline to support durable execution beyond the previous 800-second limit. Complex, long-running AI tasks now complete reliably.

Bug Fixes

  • Fixed duplicate and delete dialog issues on apps list
  • Fixed prompt references missing after connecting inputs
  • Added validation for empty prompts before generation
  • Improved error handling for GitHub Trigger configuration
  • Fixed open redirect vulnerability in OAuth callback

This release includes v0.47.0 through v0.47.5