Reads
Edge / silhouette
Free · Local · Unlimited
Shape and Blend.
These two engines run entirely in your browser — no cloud call, no credits, no limit on generations. Shape reads the silhouette of flat 2D artwork; Blend layers a grayscale-luminance read of shading and texture on top of that same silhouette. Neither one is built for a real photograph — Depth2Beam detects those on upload and points you to the Relief Generator instead (see below).
Engine 01
Shape
Built for flat 2D illustrations rather than photographs — logos, crests, line art, clean graphic designs. Shape uses a distance-transform approach: it reads the silhouette of your artwork and carves outward from the edges, rather than trying to interpret grayscale brightness as depth. That makes it the right choice whenever the source has sharp, readable shapes rather than continuous tone.
Best source
Logos, crests, line art
Cost
Free, unlimited
Reads
Silhouette + grayscale luminance
Key controls
Blend Detail Weight · Detail Strength · Denoise · Luminance Curve
Best source
Logos/crests with shading, texture, or engraved-style detail
Cost
Free, unlimited
Engine 02
Blend
Shape supplies the broad volume — the dome carved from your artwork's silhouette — while a grayscale-luminance read of the same artwork rides on top and tapers off toward the edges. Detail Strength controls how much of that fine tonal texture survives into the height map, Denoise smooths sensor/compression noise before it becomes surface roughness, and the Luminance Curve reshapes how brightness maps to depth so midtones can be pushed up or down independently of the extremes. This is the engine for artwork that's mostly flat graphic but carries real shading or engraved-style detail on top — a crest with modeled highlights, a badge with gradient fill, linework combined with shaded fill.
Already have a photograph in hand?
Shape and Blend are both built for graphic artwork — logos, crests, shaded illustrations — not real photographs. Depth2Beam detects a genuine photo on upload and won't run it through either engine; instead, describe the subject in the Relief Generator: it sculpts a relief from a text prompt and can automatically refine it with the same cloud AI Depth model, no photo upload required.
See the Relief Generator →