Head to head: Wan vs Ideogram V4.5 Edit
Wan vs Ideogram V4.5 Edit
By Ryan Merket · Published
Eight prompts put Wan and Ideogram V4.5 Edit through tests of exact text, object counts, composition, style, and negative instructions. The comparison rewards not just attractive images, but how closely each model follows the brief.
Wan’s advantage came from reliably honoring specific constraints. It rendered the requested three-line text with the right wording and separator, and more often delivered the requested 16:9 framing and scene details—in the midnight kettle, corner-store launch, tabletop arrangement, and juniper tea poster tasks, Ideogram missed or obscured key elements or shifted to a square composition. Wan wasn’t flawless: its lip-balm image appeared to contain 12 tubes, with one cropped, and it still put a plant on the windowsill in the no-plants nook prompt. But Ideogram’s corresponding image included multiple forbidden elements, while its balm scene had far too many tubes and extra objects. On these briefs, Wan’s imperfect compliance was still materially closer. The ukiyo-e fishing-boat task was the lone tie. Ideogram made the boat’s ride beneath the wave clearer and more dramatic; Wan better matched the muted indigo-and-cream palette, but muddied the scene with partly obscured boats. That result keeps the verdict grounded: Ideogram had a real strength, just not enough to change the overall pattern. Across the eight tasks, Wan scored 63.9 to Ideogram’s 46.3, winning seven with no losses; the statistical verdict is decisive at limited confidence. **Call: Wan wins—not because every image is exact, but because it follows more of the brief, more consistently.**
Legible multi-line text
A minimalist event poster with three lines of crisp, correctly-spelled text stacked and centered: 'NIGHT MARKET' large on top, 'Fridays · 6–11pm' in the middle, 'Riverside Pier 4' at the bottom, on a deep navy background, clean sans-serif, subtle grain.
OpenAI: GPT-6 Luna: Model A renders the requested three lines with the correct wording and punctuation, centered cleanly on a deep navy background with subtle grain. Model B is visually strong, but breaks the title into two lines, substitutes an asterisk for the middle dot, and uses heavier texture than requested. (Second judge pass, order swapped — scores are the average of both: Model A matches the requested three-line wording, capitalization, and separator, with a clean centered layout on a deep navy background. | Anthropic: Claude Fable 5.1: Model A renders all three lines exactly as spec…
Eleven lip balms
A top-down product photo of EXACTLY 11 distinct lip balm tubes scattered on a pale terrazzo bathroom counter, each fully visible and individually countable, with varied cap colors and label designs but all clearly separate, no boxes, no duplicates hidden behind others, no extra cosmetic items except a single small water droplet near one tube; bright diffused morning light, soft realistic shadows, sharp focus, clean commercial photography, 16:9.
OpenAI: GPT-6 Luna: Model A is closer to the requested wide terrazzo product photo, though it appears to show 12 tubes and one is cropped. Model B has many more than 11 tubes, includes extra objects and droplets, and uses a square framing rather than the requested 16:9 composition. (Second judge pass, order swapped — scores are the average of both: Model A better matches the top-down terrazzo product-photo look and 16:9 framing, with clearer separation and more polished lighting. Neither image shows exactly 11 full | Anthropic: Claude Fable 5.1: Model A delivers a true top-down 16:9 shot on a…
Midnight kettle realism
A photorealistic 16:9 kitchen counter scene at 5:42 a.m., centered on a brushed stainless steel gooseneck electric kettle with tiny condensation beads beside a walnut pour-over stand, a folded oatmeal-linen dish towel, and a hand with natural skin texture reaching in from the right; cool blue predawn window light mixes with a single warm under-cabinet strip, casting believable soft shadows, faint reflections in the metal, realistic wood grain, matte ceramic, woven fabric fibers, and subtle steam rising from a just-rinsed glass dripper.
OpenAI: GPT-6 Luna: Model A more closely matches the requested scene, with visible condensation on a convincing gooseneck kettle, a glass dripper on a wooden stand, and the cool-window/warm-strip lighting. Model B has a strong composition and includes a readable time, but its unusually ribbed kettle and missing glass dripper make it less faithful. (Second judge pass, order swapped — scores are the average of both: Model A better captures the requested predawn scene, with a condensation-dotted gooseneck kettle, visi | Anthropic: Claude Fable 5.1: Model A delivers nearly every requested element…
Corner store launch table
A lively multi-subject 16:9 scene inside a tiny neighborhood corner store during an in-store product tasting: a barista in a rust apron pours sparkling yuzu soda into sample cups, a cyclist in a silver rain jacket reaches for a chilled can, a shop owner stands behind a narrow display of artisanal snacks, and a child with a transparent umbrella points at a striped cookie tin near the register; baskets of fruit, handwritten price tags, a humming glass-door cooler, and hanging paper lanterns create depth, with all figures interacting naturally in one coherent, well-balanced composition under mixed late-afternoon window light and warm interior bulbs.
OpenAI: GPT-6 Luna: Model A captures nearly all requested details in a coherent, lively 16:9 store scene, with the tasting, silver-jacketed cyclist, shop owner, child, fruit, cooler, and lanterns clearly visible. Model B has a strong tasting interaction and several matching props, but its square framing, obscured child, and missing shop owner make it less complete and balanced. (Second judge pass, order swapped — scores are the average of both: Model A better matches the requested 16:9 scene and brings the tasting, | Anthropic: Claude Fable 5.1: Model A delivers the requested 16:9 scene with…
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 B gives the fishing boat a clearer, more convincing ride beneath a dramatic cresting wave, with a strong ukiyo-e look and a distinct dawn sun. Model A has a closer muted indigo-and-cream palette, but its multiple partly obscured boats make the scene less focused. (Second judge pass, order swapped — scores are the average of both: Model B more directly fulfills the prompt with one clearly depicted fishing boat riding a dramatic cresting wave, a dawn sun, and convincing ukiyo-e linework and | Anthropic: Claude Fable 5.1: Model A more faithfully captures the flat color…
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 more closely matches the photorealistic tabletop brief and 16:9 format, with the requested colors and object positions clearly rendered. Model B includes the right objects and placement, but looks more like polished CGI and uses a square composition. (Second judge pass, order swapped — scores are the average of both: Model A more closely matches the requested 16:9 tabletop scene while clearly placing the green cube left of the glossy red sphere, the blue cylinder behind them, and the yel | Anthropic: Claude Fable 5.1: Model A delivers the requested 16:9 frame with…
Juniper tea tin poster
A clean stylized illustration in the look of a vintage travel poster advertising a fictional everyday consumer product, the “Juniper Vale Night Tea” tin, shown on a small balcony table overlooking a foggy harbor at dusk; flat vector shapes, limited palette of deep teal, muted coral, cream, and gold, elegant simplified linework, crisp edges, artful negative space, and a calm wistful mood, with the cylindrical tea tin, porcelain cup, folded newspaper, and distant ferry lights arranged in a balanced poster-like composition, 16:9.
OpenAI: GPT-6 Luna: Model A more fully captures the vintage travel-poster brief with a balanced 16:9 layout, crisp flat shapes, decorative framing, and clearly rendered product name. Model B has an appealing harbor scene and strong product focus, but its square format and less poster-like presentation make it a weaker match. (Second judge pass, order swapped — scores are the average of both: Model A more fully delivers the vintage travel-poster look, with a balanced 16:9 layout, clear product labeling, and all requ | Anthropic: Claude Fable 5.1: Model A delivers a true 16:9 vintage travel pos…
Negation
A cozy reading nook with an armchair, a stack of books, and a mug of tea by a window — with absolutely NO plants, NO lamps, and NO artwork or picture frames anywhere in the frame. Warm afternoon light, 16:9.
OpenAI: GPT-6 Luna: Model A includes the requested nook elements and warm window light, but it still has a plant on the windowsill. Model B is visually appealing, but prominently includes plants, a lamp, and framed artwork, and it is not 16:9. (Second judge pass, order swapped — scores are the average of both: Model B has a polished, cozy scene but prominently includes multiple plants, a lamp, and framed artwork, violating all three exclusions. Model A includes a plant on the windowsill, but avoids lamps and artwor | Anthropic: Claude Fable 5.1: Model A honors the 16:9 ratio and excludes lamp…
Matchup powered by OpenRouter.