Head to head: Ideogram V4.5 Edit vs Wan
Ideogram V4.5 Edit vs Wan
By Ryan Merket · Published
Ideogram V4.5 Edit and Wan face off across eight image-editing tasks, from tightly controlled palettes to elaborate, cinematic scenes. The comparison turns on how well each model balances prompt fidelity, composition, and visual detail.
Ideogram V4.5 Edit takes the saved verdict with a **slight edge (not statistical)**: its aggregate score is **64.2**, against Wan’s **55.2**. It won five tasks to Wan’s three, with no ties. The assessment applies to this saved task sample, not to image editing in general. Ideogram’s advantage showed most in scenes with several demands to satisfy at once. It handled the causeway’s foreground pilgrims and centered one-point perspective, built out the moonlit Heliotrope Circuit brief, and made the Memphis mecha regatta feel like an active race rather than a lineup of craft. Its salvage skiff also better held onto the requested ribcage-focused, golden-hour mood. Wan had the better results on restricted palette, hands and anatomy, and attribute binding. In the latter two, the comparisons favored Wan’s clearer hand grip and more convincing tabletop materials and composition. Those wins matter: Ideogram’s edge is not a sweep, and Wan was stronger on several more controlled or object-focused briefs. **Final call: Ideogram V4.5 Edit, by a slight edge in this task sample.** Its stronger results on the more composition-heavy prompts outweighed Wan’s three wins, but the record supports a close editorial verdict—not a claim of broad superiority.
Nine-gate causeway perspective
From the exact centerline of the basalt causeway leading into the Nine Gates of Varkesh, depict a colossal obsidian citadel at the far vanishing point while, in the foreground, two human pilgrims, a pack lizard loaded with copper jars, waist-high mile markers, and towering 40-meter gate pylons all scale correctly against one another; hard-edged science-fantasy matte painting, strict one-point perspective, no warped architecture, receding lanterns and flagstones diminishing consistently into depth, 16:9.
OpenAI: GPT-6 Luna: Model A better fulfills the foreground grouping with two pilgrims and a jar-laden pack lizard, while keeping the causeway and citadel centered in clear one-point perspective. Model B has more intricate fortress detail and imposing pylons, but only one pilgrim is prominent in the foreground and its framing is less faithful to the requested format. (Second judge pass, order swapped — scores are the average of both: Model B more clearly delivers the towering repeated gate pylons, colossal citadel, | Anthropic: Claude Fable 5.1: Model A delivers a clean one-point perspective m…
Named art style
A ukiyo-e woodblock print of a fishing boat riding a large cresting wave at dawn, faithful to the flat color planes, bold outlines, and stylized foam of the tradition, muted indigo and cream palette.
OpenAI: GPT-6 Luna: Model A more closely matches the requested muted indigo-and-cream ukiyo-e style, with bold outlines, stylized foam, and a fishing boat visibly caught in the cresting wave. Model B clearly depicts dawn, but its overlapping sun discs and less subdued palette make it a little less faithful overall. (Second judge pass, order swapped — scores are the average of both: Model A has a more convincing ukiyo-e woodblock-print look, with a dramatic cresting wave, stylized foam, and a fishing boat integrated | Anthropic: Claude Fable 5.1: Model A delivers a convincing ukiyo-e print wit…
Restricted palette
A flat-vector illustration of a mountain campsite at dusk using ONLY four colors — cream, burnt orange, teal, and charcoal — with no gradients and no stray off-palette colors, balanced composition, 16:9.
OpenAI: GPT-6 Luna: Model A follows the four-color restriction more closely and presents a clear, balanced campsite-at-dusk scene with crisp flat-vector shapes. Model B includes more gear and detail, but its busier composition and apparent extra color variations make it less faithful to the restricted-palette brief. (Second judge pass, order swapped — scores are the average of both: Model A presents a cleaner, more balanced flat-vector campsite scene with a clear four-color palette and no apparent gradients. Model | Anthropic: Claude Fable 5.1: Model A delivers a clean, balanced flat-vector d…
Hands & anatomy
A natural candid photo of a person tying a friendship bracelet, both hands clearly visible with anatomically correct fingers and a plausible grip on the threads, soft daylight, shallow depth of field, 16:9.
OpenAI: GPT-6 Luna: Model A has a warm candid feel and clearly shows the bracelet-making process, though the fingers and grip look slightly less natural. Model B presents both hands and their grip more clearly with stronger anatomical definition, although it reads a little more like handling a bracelet than actively tying one. (Second judge pass, order swapped — scores are the average of both: Model B more directly matches the candid daylight prompt, with both hands clearly visible and a believable pinch on the bra | Anthropic: Claude Fable 5.1: Model B shows two clean, anatomically convincin…
Heliotrope Circuit palette
A moon-orbiting botanical lab grown inside a transparent dodecahedron greenhouse, where a bioluminescent astronomer tends spiral vines and floating seed tablets; render it as a crisp retro-futurist poster using only the named palette 'Heliotrope Circuit': heliotrope purple, acid mint, midnight navy, soft ivory, and muted charcoal, with absolutely no other colors anywhere in the image, clean graphic shapes, balanced centered composition, 16:9.
OpenAI: GPT-6 Luna: Model A more fully depicts the moon-orbiting transparent dodecahedron, glowing astronomer, spiral vines, and floating seed tablets while using the requested palette. Model B has a crisp graphic style, but its narrow poster leaves much of the 16:9 frame empty and the lunar setting is missing. (Second judge pass, order swapped — scores are the average of both: Model A more fully captures the moon-orbiting greenhouse, floating seed tablets, and crisp retro-futurist poster brief, with a stronger 16: | Anthropic: Claude Fable 5.1: Model A delivers a true 16:9 retro-futurist pos…
Memphis mecha regatta
A regatta of five amphibious duel-mechs skimming across the rain reservoir of Sector Nera during a civic festival, crowds watching from zigzag terraces and inflatable arches; depict it faithfully in named Memphis design style with flat geometric forms, squiggles, bold asymmetry, playful 1980s surface patterns, simplified figures, and bright graphic contrasts, poster-like composition under a clear afternoon sky, 16:9.
OpenAI: GPT-6 Luna: Model A more clearly depicts five-ish amphibious mechs racing across the reservoir, with energetic wakes, spectators on steep terraces, inflatable arches, and a strong festival-poster feel; Model B has especially clear Memphis patterns but its mechs read more like stationary, boxy boats and the race feels less dynamic. Model A’s extra numbered craft slightly conflicts with the requested five, while Model B adds more incidental text than the prompt calls for. (Second judge pass, order swapped — s | Anthropic: Claude Fable 5.1: Model B renders a far more faithful Memphis des…
Attribute binding
A photorealistic tabletop scene: a matte green cube to the LEFT of a glossy red sphere, a small blue cylinder BEHIND both, and a single yellow rubber duck sitting ON TOP of the cube. Colors must stay on the correct objects. Soft studio light, 16:9.
OpenAI: GPT-6 Luna: Model A follows the object placement and colors well, but Model B presents a more convincing photorealistic tabletop scene with stronger material detail and a tighter composition. Both show the duck atop the green cube and the blue cylinder behind the main objects. (Second judge pass, order swapped — scores are the average of both: Model B more convincingly matches the requested material finishes and tabletop scene, with all four objects clearly visible and correctly positioned. Model A follows | Anthropic: Claude Fable 5.1: Model A nails the spatial layout and the 'small'…
Golden-hour salvage skiff
At golden hour on the copper-salt flats of Ilyr Basin, a one-person salvage skiff with three spidery landing legs glides beside the half-buried ribcage of a colossal extinct leviathan machine, its brass cables and dusty sensor dishes glowing in low sun; cinematic 1980s retro print sci-fi illustration, long shadows stretching toward the viewer, warm amber rim light, a wistful frontier mood, wide composition 16:9.
OpenAI: GPT-6 Luna: Model A better captures the half-buried mechanical ribcage, glowing cables and sensor dishes, and wistful golden-hour frontier mood in a cinematic retro-print style. Model B has a clear, appealing composition, but its prominent skeletal skull and crisp comic-book rendering depart from the requested ribcage-focused scene and painterly look. (Second judge pass, order swapped — scores are the average of both: Model A better captures the half-buried ribcage, brass cabling, sensor dishes, lone pilot, | Anthropic: Claude Fable 5.1: Model A delivers a cinematic, half-buried colos…
Matchup powered by OpenRouter.