Head to head: Ideogram V4.5 Edit vs Kling Image
Ideogram V4.5 Edit vs Kling Image
By Ryan Merket · Published
Ideogram V4.5 Edit and Kling Image face off across cinematic scenes, layout and perspective prompts, exact counting, and constraint-following. The comparison looks at both visual execution and how closely each model sticks to the brief.
Ideogram’s strongest results came in the Storm Oracle portrait, Amber Docks sunrise, and spatial-layout task: it better captured the rain-lashed oracle clutching her ring, the open peach-and-amber harbor, and the requested room arrangement. Kling took exact counting, perspective and scale, and Gravbike Alley Burst, with a clean row of seven cups, more disciplined library vanishing lines, and a more kinetic mid-turn action scene. Kharad Loom Detail and Negation were ties. The aggregate scores were **62.6 for Ideogram V4.5 Edit** and **59.4 for Kling Image**, but the saved result remains a tie: each model won three tasks, with two ties. The aggregate difference does not change that task-by-task verdict. **Final call: Too close to call (not statistical).** This assessment applies to the saved task sample; it is not a claim about either model’s general capability.
Storm Oracle Terror
A close, waist-up cinematic portrait of Irid Vale, a young storm oracle on the glass parapet of the weather-spire of Nhalis, convincingly showing sudden, unmistakable terror as she sees something impossible in the clouds—eyes wide and wet, lower lip tensed, nostrils flared, neck tendons visible, one hand clutching a copper divination ring so hard her knuckles blanch; wind-whipped silver-blue hair, rain-specked skin, indigo ceremonial coat with tiny lightning embroidery, painted in hyper-detailed realistic fantasy concept art, with a shallow depth of field, off-center composition, stormlight and pale lightning illuminating her face against a dark teal sky, 16:9
OpenAI: GPT-6 Luna: Model A better captures the young oracle’s wind-whipped, rain-specked appearance and clutches the copper ring close to her chest, with a strong off-center cinematic composition. Model B conveys distress more forcefully through her expression and visible neck tendons, but the ring is merely held against the parapet and the figure feels less like the requested close portrait. (Second judge pass, order swapped — scores are the average of both: Model B more convincingly conveys sudden terror through | Anthropic: Claude Fable 5.1: Model A delivers a polished, cinematic concept-…
Kharad Loom Detail
An ultra-detailed studio product photograph of a folded ceremonial battle-cloak from the Kharad Ninth Fleet laid on black basalt, emphasizing fine texture and detail: dense charcoal wool weave, iridescent beetle-silk lining, hand-stitched bronze sigils, tiny soot grains caught in the seams, frayed edges repaired with crimson thread, and six matte titanium clasps etched with microscopic star charts; crisp focus across the fabric surface, controlled side lighting revealing every fiber and emboss, neutral dark backdrop, premium editorial composition, 16:9
OpenAI: GPT-6 Luna: Model A more convincingly presents a neatly folded charcoal-wool cloak as a premium studio product photograph, with visible iridescent lining, repaired crimson stitching, and crisp surface texture. Model B makes the bronze sigils and chart-like clasps more prominent, but the cloak is less densely folded and has fewer than the requested six clasps. (Second judge pass, order swapped — scores are the average of both: Model B more clearly shows the iridescent lining, crimson repairs, ornate stitchin | Anthropic: Claude Fable 5.1: Model A delivers a convincing studio product sh…
Amber Docks Sunrise
A wide cinematic sci-fi harbor scene at golden hour: the levitating cargo docks of Pelere Station just after sunrise, long amber light raking across oxidized hulls and suspended mist, casting elongated shadows from crane-arms and tether cables while the upper clouds glow peach and apricot; a lone dock inspector in a saffron pressure coat stands on a catwalk in the lower third for scale, mood quiet and expectant, rendered as painterly yet realistic science-fantasy key art with atmospheric perspective and a clear sense of early morning warmth, 16:9
OpenAI: GPT-6 Luna: Model A better captures the peach-and-amber sunrise, suspended mist, tethered levitating docks, and painterly science-fantasy key-art mood, with the inspector clearly providing scale. Model B has a compelling industrial harbor and a well-placed figure, but its oversized, dark cargo structures make the scene feel less open and its warm atmospheric sky is less prominent. (Second judge pass, order swapped — scores are the average of both: Model A more fully captures the painterly science-fantasy ke | Anthropic: Claude Fable 5.1: Model A delivers the painterly science-fantasy…
Spatial layout
A clean isometric illustration of a bedroom: a bed against the LEFT wall, a round rug centered on the floor, a desk under the WINDOW on the back wall, and a floor lamp in the FRONT-RIGHT corner. Flat-vector style, consistent perspective.
OpenAI: GPT-6 Luna: Model A clearly places the bed along the left wall, the desk beneath the back-wall window, the round rug near the center, and the lamp toward the front-right, with a polished flat-vector look. Model B includes the requested furnishings and an isometric room, but its stark monochrome rendering is less aligned with the requested style and the lamp reads as positioned near the back-right rather than the front-right corner. (Second judge pass, order swapped — scores are the average of both: Model B | Anthropic: Claude Fable 5.1: Model A delivers a clean flat-vector room with t…
Exact counting
A clean overhead flat lay on white marble of EXACTLY seven identical espresso cups arranged in a single straight row, each clearly separate and individually countable, even soft lighting, 16:9.
OpenAI: GPT-6 Luna: Model B presents exactly seven clearly separated, matching espresso cups in a straight row, with consistent orientation and clean, soft lighting. Model A also has seven cups, but the spoon crossing one cup adds an unrequested distraction and makes the arrangement less uniform. (Second judge pass, order swapped — scores are the average of both: Model B presents exactly seven clearly separated, matching espresso cups in a straight row on white marble, with clean, even lighting. Model A also has se | Anthropic: Claude Fable 5.1: Model A shows exactly seven cleanly separated,…
Perspective & scale
A photorealistic one-point-perspective shot down a long empty library aisle, shelves receding to a single vanishing point, books getting consistently smaller with distance and nothing warped or wrongly sized, warm overhead light, 16:9.
OpenAI: GPT-6 Luna: Model B has the stronger, more precisely centered one-point perspective, with shelves and books receding consistently toward a clear vanishing point and an unobstructed aisle. Model A is warm and convincing, but the ladder in the aisle and slightly less regular perspective make it a weaker match. (Second judge pass, order swapped — scores are the average of both: Model B more precisely delivers a centered one-point perspective, with shelves and books consistently diminishing toward a clear vanis | Anthropic: Claude Fable 5.1: Model A delivers a convincingly photorealistic…
Gravbike Alley Burst
A dynamic action shot of courier Jexa Morren blasting a cobalt gravbike through the flooded alleyways of Sector Veyr-12, captured mid-turn as the bike skims inches above the water, throwing a crescent spray and scattering neon reflections; her coat snaps backward, one boot lifted, gloved hand yanking the handlebars while three delivery canisters swing on magnetic tethers, with motion blur in the background signage and razor-sharp focus on her determined face and the bike chassis, composed like a high-end graphic novel cover under magenta-and-cyan night lighting, 16:9
OpenAI: GPT-6 Luna: Model B captures the sharper mid-turn action, lifted boot, dramatic crescent spray, and tethered cargo with a bold graphic-novel look, though it shows only two canisters. Model A is highly polished and includes three canisters, but the bike feels more grounded and the action less kinetic. (Second judge pass, order swapped — scores are the average of both: Model B delivers the more forceful mid-turn action, with a vivid crescent spray, a lifted boot, and a sharp graphic-novel look, though it show | Anthropic: Claude Fable 5.1: Model B delivers the requested dynamic mid-turn…
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 B creates a more intimate, warmly lit reading nook, with the chair, books, steaming tea, and window arranged as a cohesive focal scene. Model A also follows the key constraints, but its wider composition feels less focused and the books and throw on the floor make the nook slightly less inviting. (Second judge pass, order swapped — scores are the average of both: Model B more directly and cleanly fulfills the requested nook, with the books and steaming tea arranged together in vivid warm a | Anthropic: Claude Fable 5.1: Both outputs correctly omit plants, lamps, and…
Matchup powered by OpenRouter.