Head to head: Kling Image vs Ideogram V4.5 Edit
Kling Image vs Ideogram V4.5 Edit
By Ryan Merket · Published
Eight challenges put these models through cinematic scenes, strict exclusions, spatial layouts, text rendering, and detail-heavy illustration. The comparison turns on a practical question: which model follows the whole brief, not just its visual style?
Kling Image wins this matchup decisively: it scored 60.6 to Ideogram V4.5 Edit’s 51.0 and took all eight task verdicts, with 100% statistical confidence. The edge came from consistency across very different prompts, especially when composition and explicit constraints mattered as much as polish. Kling repeatedly honored the requested framing and arrangement. It delivered the wide cinematic Arctic ferry scene, put the desk beneath the window, and kept the salt-flats poster’s towers, cyclist, causeway, and reflections in the requested 16:9 layout. In the reading-nook test, it also avoided the forbidden plants, lamp, and artwork—elements Ideogram prominently added. That doesn’t make Kling flawless. Its restricted-palette campsite wandered beyond the specified colors, and its market sign appears to substitute a dot for the requested dash. Ideogram had real bright spots: its ferryman’s expression conveyed exhaustion more clearly, its time text was more accurate, and its flat-vector bedroom style was stronger. But those advantages couldn’t offset repeated misses on format, placement, and exclusions. **The call: Kling Image wins, 8–0. It is the more dependable choice when a prompt contains several concrete requirements that all need to survive the image-making process.**
Storm-Relief Ferryman
A weathered ferry pilot on the tiny Arctic island harbor of Qeqertaq Sioraq stands alone on a wet wooden dock just after a rescue, visibly conveying overwhelming relief mixed with exhaustion; his face must unmistakably show trembling gratitude—glassy eyes, loosened jaw, half-laugh half-cry expression, salt-streaked cheeks, shoulders dropping as tension releases—while he clutches a coil of orange rope against a dark green wool sweater. Render as a realistic cinematic still with cold blue dawn light breaking through storm clouds, gulls and a battered mail boat behind him, shallow sea mist, off-center composition, 16:9.
OpenAI: GPT-6 Luna: Model A better captures the stormy, misty Arctic-dawn setting and the requested 16:9 off-center cinematic composition, with the pilot visibly clutching the rope in a dark green sweater. Model B shows his face more clearly, but its square framing and calmer, brighter harbor feel less faithful to the prompt. (Second judge pass, order swapped — scores are the average of both: Model B more clearly conveys glassy-eyed relief and exhaustion, with the orange rope, green sweater, gulls, boat, and wet do | Anthropic: Claude Fable 5.1: Model A delivers the required 16:9 cinematic fr…
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 armchair, book stack, tea, window, and warm afternoon light while avoiding plants, lamps, and artwork; its wide composition also matches the requested format. Model B is attractive but prominently includes all three prohibited elements and is not 16:9. (Second judge pass, order swapped — scores are the average of both: Model B has a cozy, polished reading nook, but prominently includes plants, a lamp, and framed artwork despite the explicit exclusions. Model A incl | Anthropic: Claude Fable 5.1: Model A correctly obeys every negation (no plan…
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 better matches the requested 16:9 format and presents a clear campsite scene, though its gradient sky and extra color variations violate the strict palette rule. Model B is more elaborate and visually polished, but its square format and broader range of colors depart further from the constraints. (Second judge pass, order swapped — scores are the average of both: Model B delivers a detailed, polished flat-vector campsite scene with a mostly restricted palette, though it is square rather | Anthropic: Claude Fable 5.1: Model A delivers the 16:9 format and a mostly re…
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 places the desk directly beneath the window and follows the requested furniture layout, though its rendering is more softly shaded and 3D than flat-vector. Model B better matches the flat-vector style, but its desk is beside rather than under the window and its rug is partly covered by the bed. (Second judge pass, order swapped — scores are the average of both: Model A places the bed along the left wall, the desk directly under the window, the rug near the floor’s center, and the lamp at | Anthropic: Claude Fable 5.1: Model A places the bed on the left wall, a clea…
Monsoon Platform Reunion
A bustling multi-subject travel scene on platform 3 of the rain-soaked hill railway station at Neral Ghat during monsoon twilight: seven interacting subjects arranged in one coherent composition—a tea vendor pouring from a dented brass kettle, a violinist busker under a striped umbrella, twin backpackers unfolding a creased map, an elderly stationmaster lifting a green lantern, a child in red gumboots chasing a paper boat through a puddle, and a sleepy mountain dog beneath a bench. Render as a detailed watercolor illustration with glowing station lamps, receding tracks, drifting steam, balanced triangular grouping across the frame, layered depth, and lively but readable storytelling, 16:9.
OpenAI: GPT-6 Luna: Model A better captures the bustling, warm watercolor storytelling, with clear tea pouring, a paper boat, backpackers, and a richly layered rainy station; Model B more clearly shows the stationmaster lifting a lantern but misses the requested twin-backpacker grouping and feels more like a crisp graphic illustration. Both place the dog away from beneath the bench, though Model A comes closer to the overall mood and composition. (Second judge pass, order swapped — scores are the average of both: M | Anthropic: Claude Fable 5.1: Model A honors the 16:9 format, watercolor look…
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 better matches the requested three-line centered minimalist layout, though its middle line uses a dot instead of the dash in the specified time. Model B renders the time more accurately, but splits the title across two lines and uses a rougher, less minimalist background treatment. (Second judge pass, order swapped — scores are the average of both: Model A keeps “NIGHT MARKET” together as the prominent top line and renders the location correctly, though its middle line appears to use a d | Anthropic: Claude Fable 5.1: Model A delivers the requested three stacked, m…
Saffron-Teal Salt Flats
An expansive travel landscape of the remote Kharzul salt flats at sunset, rendered as a crisp graphic poster illustration in a strictly limited named palette: saffron, teal, ivory, charcoal, and dusty rose only—no other colors anywhere. Show mirrorlike salt pools, three stilt watchtowers, a narrow causeway with a lone cyclist, and layered clouds reflected in the water; the composition should feel elegant and minimal, with clean shapes, long shadows, and precise palette discipline across sky, land, reflections, and tiny details, 16:9.
OpenAI: GPT-6 Luna: Model A closely matches the requested 16:9 graphic-poster style and includes the three watchtowers, lone cyclist, causeway, mirrorlike pools, and reflected clouds in a disciplined palette. Model B includes the main scene elements and has appealing reflections, but its square format and more painterly rendering depart from the prompt. (Second judge pass, order swapped — scores are the average of both: Model A better matches the requested 16:9 crisp graphic-poster style and clearly includes the sa | Anthropic: Claude Fable 5.1: Model A delivers the requested 16:9 crisp graph…
Macro Compass Frost
A macro close-up photograph of an antique brass hiking compass half-buried in rime frost on the basalt rim of Mount Velin, shot with a true macro lens effect so that ice crystals on the engraved degree marks are razor-sharp while the distant alpine valley dissolves into soft blur. Early blue-hour light glints across the metal, tiny frozen lichen clings to the stone, and the framing is intimate and tactile with extreme foreground detail, narrow depth of field, and naturalistic photographic realism.
OpenAI: GPT-6 Luna: Model A more clearly captures the antique compass, sharply detailed frost, basalt, lichen, and softly blurred alpine valley, with stronger visibility of the engraved dial. Model B has appealing blue-hour atmosphere and a convincing shallow depth of field, but the compass markings are less crisp and the subject is darker. (Second judge pass, order swapped — scores are the average of both: Model A renders the frost crystals, engraved compass face, basalt, and lichen with stronger tactile detail wh | Anthropic: Claude Fable 5.1: Model A delivers crisper macro detail with razo…
Matchup powered by OpenRouter.