Arquitecto Virtual GPT-AI architectural concept generator

Example

A couple wants a 250 m² hillside home on a 15% slope, warm-temperate climate, with a $350–450k build target. They upload three precedent projects showing board-formed concrete, vertical larch cladding, and deep eaves. Arquitecto Virtual GPT: • Extracts hard constraints: slope %, approximate cut/fill sensitivity, driveway access high on the site, east-west views, 2-car parking. • Parses precedents to infer a palette (board-formed concrete base, warm timber upper volume), façade rhythm (2.4–2.7 m bay spacing), and passive strategies (breeze paths, overhangs). • Asks targeted follow-ups: bedrooms/baths, ADU/guest needs, accessibility, workshop noise isolation, solar PV interest, water harvesting, code cues (max height, 3 m setbacks). • Produces a “Design Input Card” summarizing site/topography, adjacencies, privacy gradients, acoustic concerns, budget tier, and style cues from the examples.

Scenario

Real-world use: An architect joins a kickoff call with the client. Instead of starting from a blank page, they share the Design Input Card (clear bullet points + diagrams) that Arquitecto Virtual GPT generated. The team aligns quickly on scope, functional priorities (e.g., kids’ bedrooms clustered away from living), and style direction grounded in the client’s own examples.

Example

For the same hillside brief, Arquitecto Virtual GPT outputs: • Concept A (Creative): Split-level massing stepping with the contour; concrete retaining plinth with a timber pavilion above. Entry mid-level; day spaces up one half-level for views; bedrooms below as quiet, earth-bermed zones. Passive stack ventilation via double-height stairwell; green roof over the pavilion. Preliminary area schedule (+/−5–10%), circulation diagram, sunlight/ventilation arrows, and code notes (height envelope, slope driveway limits). • Concept B (Example-Driven): Two interlocking bars on pilotis echoing the clients’ precedents. Board-formed concrete pilotis and shear walls; larch cladding on upper bar with deep eaves matching reference proportions. Carport tucked under. Rigorous window grid derived from example bay spacing. Same area schedule and a side-by-side trade-off table (e.g., Concept A: better acoustic separation; Concept B: lighter foundations on slope, simpler drainage).

Scenario

Real-world use: A developer evaluating a narrow infill site receives two clearly contrasted schemes: (A) a courtyard-heavy solution maximizing daylight factor and cross-ventilation, (B) a street-wall solution aligned with neighborhood context images they provided. The developer compares yield, leaseable areas, privacy, constructability, and façade cost ranges before selecting a direction for schematic design.

Example

For a boutique café renovation, Arquitecto Virtual GPT outputs a complete rendering brief: • Shot list: 3 exterior (eye-level hero at dusk with warm interior glow; corner wide at 28 mm midday; signage close-up), 4 interior (barista POV at 35 mm; communal table at 50 mm; ceiling feature detail; restroom accessibility view). • Time-of-day/lighting: Golden hour exteriors; interior key light from pendant grid at 3000K + indirect cove at 2700K; window exposure target EV ~9; recommended HDRIs. • Materials/specs: Tile (muted green matte, 100×100 mm, beveled, gloss 0.1–0.2), terrazzo floor (12 mm aggregate, warm base), oak veneer (vertical grain, semi-matte), blackened steel shelving (roughness 0.4), grout color, PBR hints (normal maps for board-formed concrete, bump for tile bevels), paint codes (NCS/RAL), and fabric swatches. • Entourage: Local vegetation list, density targets, staff/customer silhouettes, accessibility markers. • DCC guidance: V-Ray and Enscape equivalents for sun intensity, tone mapping, white balance; lens focal lengths; camera heights; DOF suggestions.

Scenario

Real-world use: A 3D artist opens the brief and builds the scene without back-and-forth. The materials table includes texture naming conventions and scale references (e.g., brick 65 mm height), so mapping looks real. The client reviews near-photoreal images that faithfully reflect the selected concept and example-driven style—accelerating approvals.