Veras 5.0 is here! The latest version addresses AI rendering's one-way workflow with the new Image to 3D feature. The release also brings new render engines for image and video generation, more precise reference-image compositing, AI-driven depth of field, a rebuilt Rhino sync, and a series of interface improvements aimed at making the tool usable in any window size, including phone browsers.
Jump to a feature:
Image to 3D assets
AI rendering has operated as a one-way workflow since its inception. Designers generate a striking image from their BIM or CAD model, but that image then becomes disconnected from the design process. There is no quick and practical way to transfer design intent, geometry, materials, or objects back into the authoring environment. Veras 5.0 addresses this directly with Image to 3D, a feature that makes AI rendering bidirectional for the first time.
You can now use AI to match an image to the closest Cosmos asset—which costs no credits—then place and export it into Revit, SketchUp, or Rhino at the correct scale and position. Alternatively, you can generate a 3D asset from an image using two AI models, either Tencent’s HY3D or Microsoft’s Trellis2.
Either path allows you to generate an asset or several assets at once, which you can directly export and place into your Revit, Rhino, or SketchUp project with correct scale and coordinate alignment. Veras now lets design data flow back into the BIM authoring environment rather than terminating at a flat image.
For architects and designers, this eliminates the slow, manual rework of rebuilding geometry from scratch or sourcing generic assets to approximate a rendered idea. A concept generated during an early design pass can become an accurate, editable model in the same session, cutting the concept-to-model gap that has defined AI rendering workflows until now.
Grok Imagine render engines
Veras 5.0 adds a new render engine for image generation with multiple resolution options, along with three new video engines: Grok Imagine for prompt-to-video, image-to-video, and video editing; Grok Imagine 1.5 for animating still images; and Grok Extend for extending clips beyond their final frame.
Previously, generating a still image and a video required switching between separate engines. This release consolidates that into a single high-quality pipeline that also generates native audio with every video and supports clips up to 15 seconds, extending past the last frame instead of requiring a full restart.
Insert images (compose & edit)
The new Insert Images feature lets you drop reference photos, logos, or entourage directly onto the 2D canvas, where they can be moved, scaled, and rotated like any other sketch object. This solves a longstanding limitation in reference-based prompting: the inability to point precisely at where a reference material or detail should apply.
You can now draw directly on, or point to, a specific part of an inserted reference and pair it with a prompt to modify just that area of the image. This gives you targeted control over composition and material application, and the canvas setup saves and reloads exactly as it was left, preserving markup and reference placement between sessions.
Depth of field
Veras 5.0 introduces a real, AI-driven depth-of-field tool. You click to set a focus point, then adjust blur and focus range with natural falloff, powered by two engines: Lotus (free) and Vision Banana (15 credits). This replaces flat, artificial-looking blur with a generated depth pass, so foreground and background elements fall off the way an actual camera lens behaves. For presentation and marketing renders, this adds cinematic realism without requiring post-production compositing in a separate application.
Rhino 3D sync
The Rhino integration has been rebuilt to sync on Rhino's own render pipeline, meaning materials, textures, and block instances now display in the Veras 3D viewer exactly as they appear in the Rhino viewport. This closes the disconnect that previously left users looking at untextured, generic geometry in Veras despite a fully material-mapped model in Rhino. Blocks also remain instanced, which keeps syncs faster and lighter, and camera alignment is now correct across all unit types.
Trash & recovery
Deleted renders, images, videos, and 3D models now move into a recoverable trash instead of being permanently removed, and stay restorable for 90 days. Any item can be restored to the library with a single click, and the system automatically protects every host and version with no additional install required. This removes the risk of losing work to an accidental deletion during an active project.
3D viewer display styles
A new style menu in the 3D viewer toolbar adds three modes: Shaded, Arctic Clay, and Textured Clay, alongside brighter default lighting throughout the viewer. This makes it easier to study form and massing without existing materials interfering with the read of the geometry, and replaces the dim default lighting that previously made models hard to inspect with a more legible baseline. Whichever style is selected carries directly into what Veras renders, and the choice persists between sessions.
Show last video frame
You can now switch between a video's first and last frame using the Show Original button, and bring that last frame directly into the property pane to continue working from it. This removes the extra steps previously required to extract a video's final frame and reuse it as a new starting point, streamlining iterative video workflows.
Cleaner, more flexible UI
Veras now fits windows down to 320px wide, covering docked panes, split screens, and phone browsers. The property pane has also been reorganized to place model controls first, followed by prompt, then source, matching the order most users actually reach for them. This replaces a layout that previously broke down in narrow windows and buried frequently used controls beneath less common ones, giving users full functionality regardless of screen size or window configuration.
Try Veras 5.0
Existing Veras, Enscape, V-Ray, and Corona users can update now to start using Image to 3D and the full set of Veras 5.0 features. For architects, interior and landscape designers, and 3D artists exploring AI-powered rendering workflows for the first time, take out a free trial and get started.