Head to head: Kling Image vs Ideogram V4.5 Edit
Kling Image vs Ideogram V4.5 Edit
By Ryan Merket · Published
Kling Image and Ideogram V4.5 Edit face off across eight varied image-generation and editing tasks, from tightly specified still lifes to perspective, style, and lighting challenges.
The saved results split evenly: Kling Image won **3** tasks, Ideogram V4.5 Edit won **3**, and **2** were ties. Kling’s clearest showings were the centered, uninterrupted one-point-perspective aisle and the reflective chrome-and-glass scene; Ideogram took the ukiyo-e style and negative-prompt nook tasks, and edged the marsh still life on details including the matte cobalt jar and open pin case. The close calls show why the tally matters more than any single standout. The macaque expressions drew competing readings, while neither fox image fully resolved the Zorn-palette brief: Kling better conveyed the curled pose, Ideogram had stronger negative space. The grebe result also came with a trade-off between action clarity and a more dramatic, detailed composition. Kling’s aggregate score is **62.7** and Ideogram’s is **60.7**, but the saved result is a tie. **Final call: Too close to call (not statistical).** This assessment applies to the saved task sample.
Marsh still-life binding
A hyper-detailed studio macro photograph of a naturalist’s marsh-collection arrangement on a weathered slate table, 16:9: a matte cobalt-blue ceramic jar, a glossy saffron-yellow glass vial, a brushed copper insect pin case, a pale jade-green silk field tag tied to a reed stem, a charcoal-black velvet moth tray, and a milky translucent quartz specimen, each object clearly separate and retaining its own exact color and material with no color spill or swapping; soft north-window light, crisp shadows, shallow depth of field, documentary precision.
OpenAI: GPT-6 Luna: Model A presents all six objects clearly in a polished, tightly framed macro still life, though the jar is glossier than specified and the moth is indistinct. Model B includes convincing collection details, but the quartz is cropped and the arrangement feels less cohesive, with more distracting empty space. (Second judge pass, order swapped — scores are the average of both: Model B better matches the specified matte cobalt jar and presents the copper pin case open, with a clearly visible moth an | Anthropic: Claude Fable 5.1: Model A keeps all six objects fully in frame, c…
Storm-guilt macaque
A cinematic close wildlife portrait of a Formosan rock macaque clutching a stolen mangosteen to its chest, looking directly into camera with unmistakable guilty alarm—brows lifted and pinched, eyes widened, lips tense and slightly parted, shoulders hunched as if caught in the act—rainforest leaves framing the face, moody monsoon backlight with beads of rain on fur, shallow depth of field, documentary photojournalism realism.
OpenAI: GPT-6 Luna: Model B conveys the lifted-brow, wide-eyed guilty alarm more clearly and has a balanced, cinematic composition with convincing rainforest depth. Model A shows the rain and mangosteen especially well, but its expression reads less unmistakably alarmed and the portrait is quite dark. (Second judge pass, order swapped — scores are the average of both: Model B more clearly conveys the requested guilty alarm through its widened eyes, tense parted lips, and direct gaze, while the close framing and moo | Anthropic: Claude Fable 5.1: Model A delivers a more convincing 'caught in t…
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 more convincingly captures the flat color planes, bold outlines, stylized foam, and balanced composition of ukiyo-e, with a clear boat riding a large cresting wave. Model A has a stronger dawn glow, but its crowded, unusually fragmented wave and boats make the scene feel less cohesive. (Second judge pass, order swapped — scores are the average of both: Model B more clearly centers a fishing boat riding a single towering crest, with restrained indigo-and-cream color planes, bold outlines, | Anthropic: Claude Fable 5.1: Model A essentially reproduces Hokusai's Great…
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 A more precisely delivers a centered, empty aisle with shelves and warm overhead lights converging cleanly at a single vanishing point. Model B is also convincing, but the bright doorway and visible cart/rail distract slightly from the requested uninterrupted aisle. (Second judge pass, order swapped — scores are the average of both: Model A presents the clearest centered one-point perspective, with the shelves, lights, and books receding consistently toward a single vanishing point. Model | Anthropic: Claude Fable 5.1: Model A delivers a very clean one-point perspect…
Grebe chase burst
A documentary wildlife photograph, 16:9, capturing a western grebe mid-courtship sprint across a mountain lake, both feet visibly slapping the water as it runs upright with wings half-open, droplets flung in sharp arcs behind it, neck stretched forward, wake tearing through mirrorlike dawn reflections, distant spruce shoreline blurred by speed, low telephoto angle, cold sunrise sidelight, intense dynamic energy and convincing motion frozen at the peak instant.
OpenAI: GPT-6 Luna: Model B more clearly captures the upright water-running courtship pose with wings spread and a wake trailing behind, while Model A is more convincingly detailed and dramatic but reads more like a low flight or takeoff. Model A has the stronger close telephoto composition; Model B’s bird is small and shadowed, though its action matches the prompt better. (Second judge pass, order swapped — scores are the average of both: Model B more clearly captures the upright water-running courtship burst, wit | Anthropic: Claude Fable 5.1: Model A delivers a convincing low telephoto doc…
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 captures the requested warm afternoon light especially well and includes all the required reading-nook elements without any forbidden objects. Model B is also cozy and adheres closely, but its cooler, dimmer palette feels less like warm afternoon light. (Second judge pass, order swapped — scores are the average of both: Model B includes the requested nook elements and keeps the frame free of plants, lamps, and artwork. Model A is also attractive and well composed, but visible foliage out | Anthropic: Claude Fable 5.1: Both include the armchair, books, tea mug, and…
Zorn fox palette
A painterly portrait of an Arctic fox curled on a snow-dusted basalt ledge in the strict Zorn palette only—ivory black, titanium white, yellow ochre, and cadmium red light—with no other colors anywhere in the image; limited-palette oil painting style, overcast polar daylight, subtle tonal modeling, clean negative space, restrained composition emphasizing palette fidelity.
OpenAI: GPT-6 Luna: Model A more clearly shows a curled fox on a snow-dusted basalt ledge and stays closer to the restricted palette, though its setting is busier and some red markings look unnatural. Model B has stronger negative space and polished brushwork, but the fox is not curled and the warm lighting and blue-gray snow drift from the requested palette and overcast daylight. (Second judge pass, order swapped — scores are the average of both: Model A more clearly shows a curled fox on a snow-dusted basalt ledg | Anthropic: Claude Fable 5.1: Model A nails the brief more closely—curled fox…
Reflections & glass
A photorealistic close-up of a chrome teapot on a polished dark marble counter next to a glass of water; the window and a red apple on the counter must be correctly reflected in BOTH the chrome and the water's surface, dramatic side light, 16:9.
OpenAI: GPT-6 Luna: Model A more clearly shows the window and red apple reflected in the chrome and glass, with dramatic side light and a strongly polished marble counter. Model B is cleanly composed, but the specified reflections and dramatic lighting are much less evident. (Second judge pass, order swapped — scores are the average of both: Model A more clearly conveys the window and apple reflections across the polished chrome and water, with dramatic side light and a strong reflective composition. Model B has th | Anthropic: Claude Fable 5.1: Model A delivers the dramatic window side light…
Matchup powered by OpenRouter.