What is Veras?
Veras is an AI-powered architectural visualization plugin developed by EvolveLAB, a Colorado-based software company founded by Bill Allen and a team of architects and technologists. Unlike generic AI image generators (Midjourney, DALL-E, Stable Diffusion) that create images from text prompts alone, Veras is geometry-based: it uses your existing 3D model — built in SketchUp, Revit, Rhino, or Vectorworks — as the structural foundation, and then applies AI-driven materials, lighting, atmosphere, and entourage based on natural language prompts. This fundamental difference means Veras respects your design intent. The building's form, massing, window placement, and structural elements come from your model. The AI controls the visual presentation: the time of day, weather, camera angle, material finishes, landscaping, people, and overall mood. As of 2026, Veras is used by over 15,000 architects and designers at firms including Gensler, HOK, Perkins&Will, and hundreds of smaller practices worldwide.
Veras integrates directly into the architect's existing workflow as a plugin: you work in SketchUp, Revit, Rhino, or Vectorworks as normal. When you want to generate a rendering, you activate the Veras plugin, which captures your current 3D view, sends the geometry (not the entire model file) to Veras's cloud-based AI, and returns a rendered image within 30-90 seconds depending on complexity and resolution. You can iterate rapidly — adjust the prompt ("make it a rainy evening with warm interior lighting visible through the windows"), tweak the creativity slider (how much the AI is allowed to deviate from the base geometry), adjust the geometry influence (how strictly the AI respects the model edges), and regenerate. This iterative feedback loop — model → prompt → render → adjust → re-render — enables architects to explore dozens of visual directions for a design in the time it previously took to set up materials and lighting for a single traditional render.
Key Features
Geometry-Based AI Rendering
Veras's core innovation is its geometry-based approach. Unlike Midjourney or DALL-E, which generate images from text and have no understanding of your specific building design, Veras captures the 3D geometry from your model — walls, roofs, windows, doors, columns, slabs — and uses it as the structural constraint for the AI rendering. This means the building in the rendering is your building, not an AI hallucination of a building that happens to match your prompt. The "Geometry Influence" slider (0-100%) controls how strictly the AI adheres to the model edges. At 100%, the rendering is tightly constrained to the model geometry — window frames stay where you placed them, the roofline matches your design, structural elements are preserved. At lower settings, the AI can add organic elements (vegetation growing on walls, weathering effects, terrain blending) that soften the edges between the designed and natural environment. For concept design presentations, architects often use 70-85% geometry influence to maintain design integrity while allowing atmospheric creativity. For client presentations where accuracy is critical, 90-100% ensures the rendering faithfully represents what will be built. For massing studies and early concept exploration, 40-60% allows the AI to suggest design variations that the architect may not have considered — transforming a massing block into a specific architectural style or material palette.
Natural Language Prompting with Creative Controls
Veras uses natural language prompts to control the visual aspects of the rendering. Example prompts: "a modern glass office building at golden hour, warm sunlight reflecting off the facade, with landscaped plaza in the foreground, professional photography style" or "cabin in the woods during autumn, foggy morning, cedar siding and stone foundation, moss on the roof, cozy interior lights visible." The prompt controls: time of day and lighting conditions (golden hour, overcast, night with artificial lights, morning mist), weather and atmosphere (fog, rain, snow, clear sky, storm approaching), materials (glass, concrete, wood, steel, brick, stone — and specific finishes like "weathered corten steel" or "white oak slats"), architectural style (modernist, brutalist, vernacular, biophilic, parametric), context (urban street, suburban neighborhood, forest clearing, coastal, desert), and entourage (people, furniture, vehicles, landscaping, water features). The "Creativity" slider controls how much the AI interprets the prompt versus relying on the base image. Low creativity produces clean, conservative renders. High creativity introduces artistic interpretation, unusual lighting, and atmospheric effects that can make a rendering feel like a photograph from an architectural magazine rather than a standard visualization. The "Prompt Assist" feature (launched 2025) suggests prompt improvements based on your model's characteristics — if the AI detects a glass-heavy facade, it suggests lighting-related prompts that complement reflective surfaces.
Iterative Rendering & Design Exploration
The speed of Veras — 30-90 seconds per render — transforms rendering from a final presentation activity to a design exploration tool. In a traditional workflow, rendering is the last step: you finalize the design, spend hours or days setting up materials, lighting, and camera angles in a rendering engine (V-Ray, Enscape, Lumion), and produce a handful of images for the client presentation. With Veras, you generate renders throughout the design process. At the massing stage: block out a building form, prompt "modern residential tower, glass and aluminum facade, twilight," and generate 5 variations with different massing proportions in under 10 minutes. At the design development stage: refine the model with window placement and facade detailing, prompt different material palettes and lighting conditions, and present the client with 20 visual options for their building rather than 2-3. This speed enables "visual conversations" with clients: "You prefer the warm wood and stone palette over the all-glass look? Let me adjust the prompt and show you three variations of that in the next 5 minutes." The ability to iterate visually during client meetings — rather than "I will get back to you with renders next week" — is a competitive advantage in winning and retaining projects. Veras saves rendering history, so you can compare iterations side by side and trace the design's visual evolution.
Multi-Platform Plugin Integration
Veras is available as a plugin for the four most widely used architectural software platforms: SketchUp (the most popular, where Veras has the largest user base), Revit (for BIM workflows — Veras captures 3D views directly from Revit models), Rhino (including Grasshopper integration for parametric design workflows), and Vectorworks (for the landscape and entertainment design communities). The plugin model means architects do not leave their design environment: they click the Veras button inside their familiar software, the current view is captured, and the rendering workflow begins. The plugin captures geometry, materials (basic color assignments from the model), and camera position. For Revit users working in BIM, this means Veras can render directly from the coordinated model without exporting, converting, or re-texturing — the model that produces construction documents also produces presentation renderings. Veras runs on Windows and Mac for all supported platforms. A web-based viewer allows sharing and reviewing renders with team members who do not have Veras or the modeling software installed — useful for client reviews and remote collaboration. The web viewer supports comments and markup, enabling a structured feedback loop between the design team and clients or stakeholders.
Entourage & Context Generation
Veras can generate contextual elements that bring architectural renderings to life: people (diverse ages, activities, clothing styles appropriate to the climate and building type — business attire for office buildings, casual for residential, tourists for cultural projects), vehicles (cars, bicycles, public transit consistent with the urban or suburban context), landscaping (trees, shrubs, flowers, grass, hardscape appropriate to the climate and season), furniture and interior elements (visible through windows, giving depth to the rendering), signage and streetscape elements, and environmental effects (puddles after rain, snow accumulation, fallen leaves, reflections in wet surfaces). These contextual elements are generated by the AI — you do not need to manually place entourage components on your model. The prompt controls the style and density: "busy urban street with diverse pedestrians and passing cars" versus "quiet residential street with one person walking a dog." For client presentations, this is transformative: instead of presenting a sterile building isolated on a blank background, you present the building as it would be experienced — with life, activity, and atmosphere. For public consultation and community engagement, the ability to show a proposed building in a recognizable, lived-in context helps stakeholders understand the project's impact on the neighborhood. Veras's entourage generation respects the building's function — it will not place office workers outside a residential building or beachwear in a winter scene (most of the time — AI quirks still occur and require prompt refinement).
Style Presets & Material Exploration
Veras includes a library of style presets — pre-configured combinations of prompts, creativity settings, and geometry influence that produce specific visual styles: "Architectural Photography" (clean, professional, magazine-quality), "Watercolor Concept" (artistic, hand-rendered feel for early presentations), "Dusk Atmosphere" (dramatic lighting with artificial interior lights), "Winter Context" (snow, bare trees, warm interior glow), "Lush Landscape" (mature vegetation, integrated with the building), "Minimalist" (clean lines, neutral palette, sparse entourage), and "Aerial Site" (bird's-eye view for master planning and site context). These presets serve as starting points that architects customize with additional prompt details. For material exploration, Veras can generate the same view with different material palettes — prompt "change the facade from glass to brick and limestone" and compare side by side. This capability is particularly valuable during the design development phase when clients and architects are making material decisions that have significant cost and aesthetic implications. The material exploration feature uses the model geometry to maintain consistency — window openings, floor heights, and structural rhythm remain the same across material variations, isolating the material as the changed variable. This is far more efficient than traditional rendering workflows where changing materials requires re-texturing the model and re-rendering from scratch.
Veras Pricing (2026)
| Plan | Price | What You Get |
|---|---|---|
| Individual | $39/month ($29/month annual) | 1 user, unlimited renders, all plugins (SketchUp, Revit, Rhino, Vectorworks), 1024px output resolution, standard render queue, email support. For solo practitioners and freelancers. |
| Studio | $79/month ($59/month annual) | 1 user, unlimited renders, 2048px output resolution with upscaling, priority render queue, Prompt Assist AI, style presets library, advanced creative controls, web viewer for sharing, priority support. For architects who need presentation-quality output. |
| Firm | Custom pricing | Multiple users, 4K output resolution, team library (shared prompts, presets, renders), admin dashboard with usage analytics, SSO integration, dedicated account manager, custom model training (train Veras on your firm's rendering style for consistent output). For architecture firms with 5+ users. |
Pricing verified June 2026. Annual plans save approximately 25%. All plans include unlimited renders — there is no render credit system or per-image charge. The Individual plan is adequate for concept design and internal visualization; the Studio plan's higher resolution and advanced controls are worth the upgrade for client-facing presentation renders. EvolveLAB offers a 14-day free trial with full Studio plan access.
Pros & Cons
Pros
- Geometry-based rendering respects your actual design: This is the feature that separates Veras from generic AI image generators. The building in the rendering is your building — not an AI's interpretation of what a building might look like. The geometry influence slider gives you precise control over this relationship.
- Speed transforms rendering from final presentation to design exploration tool: 30-90 second render times enable rapid visual iteration during the design process rather than waiting until the end. Architects report generating 50-200 renders per project with Veras compared to 5-10 with traditional rendering workflows.
- Direct plugin integration with the four most-used architectural platforms: No export, no conversion, no leaving your design environment. Architects already working in SketchUp, Revit, Rhino, or Vectorworks can add Veras without changing their modeling workflow.
- Natural language prompting makes rendering accessible to non-specialists: You do not need to learn material libraries, lighting setups, or rendering engine parameters. "Brick facade at sunset with cafe seating on the sidewalk" is a prompt any architect can write. This democratizes high-quality visualization within firms — junior staff can produce presentation-ready renders.
- Style presets and prompt suggestions lower the learning curve: New users can start with presets and gradually learn to craft effective prompts. Prompt Assist provides helpful guidance without being intrusive.
Cons
- Not a replacement for production rendering (V-Ray, Enscape, Lumion): Veras is best for concept design, design exploration, and presentation renders. It is not suitable for construction-level visualization (accurate material specifications, precise lighting analysis, daylight studies for LEED certification) or for producing renders where every material and lighting condition must be exactly specified and reproducible. Most firms use Veras for early-to-mid-stage visualization and switch to traditional renderers for final construction-level visualization.
- AI can sometimes "drift" from the design intent: At lower geometry influence settings or with highly creative prompts, the AI may add or modify elements that do not exist in the model — a tree growing through a roof, extra windows, or altered proportions. The architect must review each render for design fidelity. This is generally manageable with the geometry influence slider but requires attention, especially for client-facing presentations where an unrealistic element could undermine credibility.
- Quality varies with prompt quality: Like all AI image tools, Veras responds to prompt quality. Architects must learn effective prompting techniques — being specific about materials, lighting, and atmosphere — to get consistent results. The Prompt Assist feature helps but does not eliminate the learning curve. A poorly written prompt produces mediocre renders regardless of the underlying model quality.
- Cloud-based means render times depend on server load: During peak usage periods (typically weekday afternoons in North American and European time zones), render queue times may extend from 30 seconds to 2-3 minutes. This is still fast compared to traditional rendering but can interrupt rapid iteration workflows. The Studio and Firm plans include priority queuing.
- Not a full BIM visualization solution: Veras captures geometry and basic colors from Revit models but does not integrate with BIM metadata (material schedules, cost data, sustainability parameters). For firms requiring renders tightly connected to BIM data, Veras is a supplementary visualization tool, not a replacement for BIM-native visualization workflows.
Frequently Asked Questions
How is Veras different from using Midjourney for architectural rendering?
Midjourney generates images from text prompts alone — it has no understanding of your specific building design. You can prompt "modern glass office building" and get a beautiful rendering of a generic modern glass office building, but it will not be your building. The window placement will be different, the proportions will be different, the massing will be different. For concept inspiration and mood boards, Midjourney is excellent. For rendering your actual design, it is fundamentally limited. Veras captures your 3D model geometry and renders it — the building in the rendering is the building you designed. The AI controls the visual atmosphere (lighting, materials, weather, entourage) but the structure is yours. This distinction is critical for professional architectural practice: presenting a client with a Midjourney-generated image of "a building" that does not match what will be built is misleading. Presenting a Veras-generated rendering of your actual design is a faithful visual representation. Many architects use both tools: Midjourney for early concept inspiration and mood exploration, Veras for rendering specific designs from their working models. The tools are complementary, but for any visualization intended to represent an actual project, Veras is the appropriate tool.
Does Veras work with detailed Revit models or just massing studies?
Veras works with models at any level of detail, from simple massing blocks to fully detailed BIM models with curtain wall mullions, structural framing, and interior partitions. The plugin captures the visible geometry in the current 3D view — what you see in Revit (or SketchUp/Rhino/Vectorworks) is what Veras renders. For detailed Revit models, Veras captures: wall assemblies (exterior finish, core, interior finish), windows and curtain walls (frames, mullions, glazing), roofs (form, overhangs, parapets), structural elements (columns, beams, bracing — if visible), site elements (topography, hardscape, planting — if modeled), and furniture and interior elements (if included in the model). The level of detail in the render matches the level of detail in the model. A massing study produces a massing render; a detailed construction model produces a detailed render. For best results with detailed models: use a 3D view that shows the desired level of detail (hide interior partitions if you want an exterior-only render), set the geometry influence to 85-100% to preserve fine details, and use prompts that are consistent with the model's detail level — prompting "add intricate curtain wall detailing" on a model that already has detailed curtain walls helps the AI enhance rather than replace those details.
Can Veras generate interior renderings as well as exterior?
Yes — Veras can render interior views from your 3D model. The workflow is identical: position your camera inside the model (in SketchUp, Revit, Rhino, or Vectorworks), activate Veras, and prompt the interior scene. Interior prompts might include: "warm minimalist living room, natural oak flooring, white walls, large windows with forest view, afternoon sunlight casting shadows across the floor, cozy reading chair and floor lamp," or "modern open-plan office, exposed concrete ceilings, suspended linear lighting, collaborative workstations, glass-walled meeting rooms, biophilic elements." Interior rendering is generally more demanding than exterior because the AI must maintain spatial coherence — walls, floor, ceiling, and furniture must all relate to each other correctly within the enclosed space. Veras's geometry-based approach helps significantly because the model provides the room dimensions and major furnishing placement. For best interior results: include basic furniture massing in your model (simple blocks indicating where furniture goes), maintain high geometry influence (90-100%) to keep walls and openings consistent, use prompts that describe lighting conditions (interior renders depend heavily on lighting for atmosphere), and be specific about material finishes. Interior renders typically take slightly longer (60-120 seconds) than exterior renders due to the complexity of maintaining spatial coherence in enclosed spaces. For high-end interior presentations where every material, fixture, and furniture piece must be exactly specified, traditional rendering (Enscape, V-Ray) remains the more precise tool — but for design exploration and client communication, Veras interiors are fast and effective.
How Architects Use Veras: Design Exploration Workflow
Start Early — Use Veras at Massing Stage
Many architects make the mistake of waiting until the design is "ready for rendering" before using Veras. The platform is most valuable at the earliest design stages. Block out a simple massing model in SketchUp or Revit — basic volumes representing the building footprint, height, and major forms. Activate Veras with a prompt describing the intended architectural language: "contemporary museum with cantilevered upper floors, limestone and glass facade, landscaped public plaza." The AI will interpret the massing through the lens of the prompt, suggesting material treatments, fenestration patterns, and contextual elements that the architect has not yet modeled. This early-stage rendering serves as a visual conversation starter — it helps the design team align on the architectural vision before committing to detailed modeling. It also surfaces design issues: the Veras render might reveal that the building reads as too massive from a particular angle, that the entrance is not prominent enough, or that the relationship between the building and the street needs more attention. These insights arrive days or weeks earlier in the design process than they would with traditional rendering workflows.
Present Options, Not a Single Direction
During client presentations, use Veras to show the same building massing with different material palettes, lighting conditions, and contextual treatments. Generate 3-5 variations of the same view: "all glass and steel," "brick and limestone," "timber and copper," each at the client's preferred time of day. This visual comparison gives clients a concrete basis for making material and aesthetic decisions — far more effective than showing material sample boards and asking them to imagine how the building will look. For public projects or projects requiring community engagement, Veras-generated streetscape views that show the proposed building in its existing context — with accurate scale, materials, and entourage — help non-architect stakeholders understand the project's visual impact. The ability to generate these views in minutes rather than commissioning a rendering team for days makes it practical to present genuinely multiple options rather than the traditional "here's our preferred option and one alternative."
Use Rendering History to Track Design Evolution
Veras saves your render history — every prompt, every image, every iteration. This creates a visual record of the design's development that is valuable for: internal design reviews (understanding why certain decisions were made), client communication (showing "here's where we started and here's where we are now"), and portfolio documentation (demonstrating the design exploration process for award submissions and firm marketing). At project milestones, export a selection of renders showing the design evolution — from initial massing studies through material exploration to final presentation renders. This narrative of the design journey is compelling for clients and juries alike. The render history also serves as a prompt library — successful prompts can be reused and adapted for future projects, building a firm-specific rendering knowledge base over time.