Runchat Updates: August 26, 2026
Major enhancements to Clipchat video editing, smarter PDF reading from the canvas, and improved multi-CAD screenshot targeting.
🎬 CLIPCHAT VIDEO EDITING IMPROVEMENTS
We've significantly upgraded Clipchat to give you more control over your video content. You can now generate clips directly from a folder of content on your laptop or a collection of assets on your runchat canvas to create automated video templates. The editor UX has been overhauled: you can now drag clips, insert gaps in the main track, and create multiple overlays. Clip configurations are now displayed in a sidebar instead of a popover for a more focused editing workflow. We've also made a bunch of quality of life improvements:Â
- Better handling of clip transitions: dragging clip ends near a cut now inserts a gap by default, preventing accidental shifts of neighboring content.
- Agents can now toggle the webcam between a framed video view and full bleed, with smooth interpolation between the two states.
- We’ve optimized the backend processes for video editing to reduce latency when switching to the edit view.
📖 SMARTER CANVAS PDF READING
We've improved how agents read and interpret PDF documents placed on the canvas. This update addresses common gaps in PDF analysis, making it easier for agents to extract information and provide insights from complex documents.
📸 MULTI-CAD SCREENSHOT TARGETING
For those working across multiple design platforms simultaneously, we've added better targeting for screenshots. Agents can now identify and capture specific CAD platforms even when multiple applications are open at once.
