What's New - December 2025 Update

We've been busy adding powerful new features to make your creative workflow even better. Here's everything that's new this month:

New AI Models

• SAM 3 Image Segmentation : Detect and isolate any object in your images using simple text prompts like "wheel", "person", or "car". Perfect for creating masks for further editing.

• SAM 3D Objects : Turn any image into a 3D model! Simply describe the object you want to extract (e.g., "chair", "car") and get a fully textured GLB file ready for use.

Workflow Builder Improvements

Real-Time Progress Updates

• Watch your workflows execute in real-time with live status updates for each node. See exactly what's processing and when it completes.

• Auto-Layout Button

• One-click automatic arrangement of your workflow nodes for a clean, organized canvas.

New Utility Nodes

• Video Frame Extraction - Extract frames from videos at specific timestamps

• Image Resize & Crop - Precise control over image dimensions

• AI Image Describer - Automatically generate descriptions of images

• Prompt Concatenator - Combine multiple text inputs

Visual Previews for Utility Nodes

Utility nodes now show visual previews of their output, making it easier to understand your workflow results.

Collapsible Quick Start Examples

The Quick Start Examples section on model pages is now collapsible, giving you more space for your work.

Failed Generation Feedback

Failed generations now show a clear error indicator instead of an endless loading spinner, plus detailed error messages are recorded for troubleshooting.

We're constantly working to improve your experience. Have feedback or feature requests? Let us know!