Head to head: FLUX1.1 [pro] vs Fibo Gen 1.5 Text to Image
FLUX1.1 [pro] vs Fibo Gen 1.5 Text to Image
By Ryan Merket · Published
This matchup wasn’t subtle: FLUX1.1 [pro] produced some handsome images, but Fibo Gen 1.5 Text to Image was the model that actually listened. Across prompt-heavy tests, Fibo was consistently more literal, more controlled, and far more reliable.
FLUX1.1 [pro] has obvious aesthetic instincts. It won the named-art-style test by delivering the more convincing ukiyo-e interpretation, and even in several losses it looked polished, cinematic, and commercially appealing. But this comparison wasn’t about who could make the prettiest image in isolation. It was about who could satisfy the brief. On that question, FLUX kept drifting off-script. Fibo Gen 1.5 Text to Image won **6 of 8 tasks**, with **96% confidence**, and the pattern behind those wins is more important than the raw count. It was better at constraint following: the restricted-palette vector scene stayed restricted; the Memphis smart speaker actually looked like the specified product on the specified pedestal; the one-point-perspective bookstore read with cleaner spatial logic; the friendship-bracelet hands were more task-specific and anatomically believable; and the server-corridor drone-parts action shot included the concrete prompt details FLUX glossed over. Even in the repair-bench test, where both models stumbled on the no-headset instruction, Fibo still tracked the requested top-down line-art setup much more closely. FLUX’s problem is clear from the losses. It too often substitutes vibe for compliance. In the restricted-palette task, it broke the color rules and slipped into painterly shading. In the repair-bench scene, it flatly violated the no-headphones requirement. In the Memphis prompt, it missed the object-and-pedestal brief. In the drone-parts burst, it delivered drama instead of the described mechanics. That makes FLUX feel like a model you use when you want attractive interpretation, not when you need dependable execution. The only genuinely close result was the neon-rain rooftop scene, which ended in a tie because each model captured different parts of the brief better. But one split decision doesn’t change the larger story. Fibo was the stronger prompt follower across composition, object specificity, perspective, anatomy, and stylistic constraint, while FLUX mostly survived on taste and finish. **Final call: Fibo Gen 1.5 Text to Image wins decisively. FLUX1.1 [pro] is the more seductive improviser, but Fibo is the better image model when the prompt actually matters.**
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.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/0e89ace0-ff75-4183-9454-c5fa2d2be9a4.jpg)

Model B adheres much better to the restricted-palette flat-vector brief, using clean solid shapes and essentially only the specified cream, burnt orange, teal, and charcoal in a balanced campsite composition. Model A is attractive and cinematic, but it breaks the prompt with many extra hues, painterly shading/gradient-like transitions, and a less strictly flat-vector treatment. (Second judge pass, order swapped — scores are the average of both: Model B closely matches the prompt with a flat-vector campsite scene, balanced 16:9 composition, and a tightly restricted four-color palette without obvious off-palette intrusions. Model A is visually appealing, but it breaks the core constraints with painterly rendering, gradients, and many extra colors beyond the specified palette.)
No-Headset Repair Bench
A meticulous electronics repair bench inside a compact satellite-ground-station workshop, centered on an opened teal handheld game console labeled "Kite-9" with ribbon cables, precision screwdrivers, a ceramic parts tray holding exactly seven brass screws, sticky notes with tiny circuit sketches, and a magnifying lamp; the scene must deliberately contain NO headphones, NO earbuds, and NO headset of any kind anywhere in frame; minimalist line art with selective pale cyan and amber accents, crisp top-down composition, soft diffused morning window light, 16:9.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/ed49e992-e36f-42f6-bc52-d524952615ec.jpg)

Model B adheres much more closely to the prompt with a clean top-down repair-bench scene, an opened teal handheld console, line-art styling, and no visible headphones or headset. Model A is visually rich but clearly violates the explicit no-headphones requirement and misses the minimalist line-art look, despite stronger text rendering and overall polish. (Second judge pass, order swapped — scores are the average of both: Model B follows the top-down repair-bench concept much more closely, with an opened teal handheld labeled "Kite-9," sticky notes, tools, and a parts tray that appears to contain exactly seven brass screws; its main miss is the presence of headphone-like items despite the explicit prohibition. Model A has stronger visual richness but clearly violates the prompt with multiple headphones/headset-like objects, a less minimalist line-art style, and a busier composition that departs from the specified restrained aesthetic.)
Memphis Smart Speaker
A compact smart speaker prototype called "Pebble Loop" displayed on a zigzag pedestal in bold 1980s Memphis design style, with playful squiggles, terrazzo patterns, off-kilter geometric shapes, teal-black-coral-yellow palette, dotted laminates, and exaggerated graphic shadows; faithful named-style execution, clean studio setup, front three-quarter product shot with bright even showroom lighting, 16:9.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/087443c7-8c90-4f9e-a6b2-53962f790693.jpg)

Model B adheres much more closely to the prompt with a compact smart speaker prototype on a zigzag pedestal, strong Memphis motifs, the specified teal-black-coral-yellow palette, and a clean front three-quarter studio presentation. Model A has appealing color and pattern work, but the speaker is too generic, the pedestal is not clearly zigzag, and the overall scene misses the richer off-kilter geometric styling and graphic-shadow energy requested. (Second judge pass, order swapped — scores are the average of both: Model B matches the prompt much more closely with a compact smart speaker prototype on a zigzag pedestal, strong Memphis motifs, the specified teal-black-coral-yellow palette, and a clean front three-quarter studio product shot. Model A is attractive but diverges from the requested object and setup: the speaker is a simple round disc, the pedestal is not clearly zigzag Memphis architecture, and the scene uses a more poster-like backdrop than a clean showroom presentation.)
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.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/8bf8ecc1-73d4-42fe-8b28-5656e802b716.jpg)

Model A better matches the requested ukiyo-e woodblock look with flatter color planes, a muted indigo-and-cream palette, and stylized foam while clearly depicting a fishing boat on a large cresting wave at dawn. Model B is strong and visually striking, but it feels more like a modern graphic illustration than a faithful ukiyo-e print, with heavier linework and a less convincing dawn atmosphere. (Second judge pass, order swapped — scores are the average of both: Model A more convincingly matches the named ukiyo-e tradition, with a composition and wave treatment that feel closer to classic woodblock prints while still clearly depicting a fishing boat on a cresting wave. Model B is strong and attractive, but it reads as a more modern graphic interpretation and is slightly less faithful to the traditional flat planes and stylized woodblock character requested.)
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.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/0e7c8ca5-f6bc-4a65-80cc-4081b5337641.jpg)

Model B better matches the requested one-point perspective, with shelves and floor lines converging cleanly to a single vanishing point and book scale diminishing more consistently into the distance. Model A is strong and photorealistic, but the depth cues are softer and the perspective feels slightly less precise and dramatic than Model B. (Second judge pass, order swapped — scores are the average of both: Model B better matches the prompt with a cleaner one-point perspective, stronger sense of books consistently shrinking into the distance, and warmer overhead lighting. Model A is also convincing and photorealistic, but the softer focus and less precise vanishing-point emphasis make its perspective and scale slightly less exact.)
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.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/220e8834-f069-4583-a38f-5b55c76786a6.jpg)

Model B better matches the prompt by clearly showing both hands tying a friendship bracelet with a believable grip on colorful threads, plus natural daylight and pleasing shallow depth of field. Model A is attractive and technically solid, but it reads more like holding a small knot of yarn than actively tying a friendship bracelet, and the hand pose feels less clearly task-specific. (Second judge pass, order swapped — scores are the average of both: Model B better matches the prompt by clearly showing both hands tying a colorful friendship bracelet with believable finger anatomy, soft daylight, and shallow depth of field, though it misses the requested 16:9 framing. Model A has strong composition and realism, but it looks more like tying a generic piece of yarn than a friendship bracelet and also does not match the 16:9 aspect ratio.)
Neon Rain Router Rooftop
A weatherproof experimental mesh router array on the roof of a mid-rise apartment block during neon-soaked midnight rain, with three slim antenna fins glowing magenta and electric blue, puddles reflecting a pharmacy cross and a noodle-bar sign from the alley below, wet cables snaking toward a humming battery pack, and a lone technician's transparent umbrella at the edge of frame; cinematic realism, moody neon night lighting, low-angle composition with deep perspective and mist in the background, 16:9.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/1f44371d-888d-4e46-aa5e-1408c8c21969.jpg)

Model B adheres more closely to the prompt by clearly showing exactly three slim glowing antenna fins, visible puddle reflections, wet cables leading to a distinct battery pack, and the transparent umbrella at the frame edge. Model A has strong mood and cinematic depth, but it misses the specified three-fin array, makes the umbrella-holder too central, and the battery pack/pharmacy-noodle reflections are less clearly resolved. (Second judge pass, order swapped — scores are the average of both: Model A adheres more closely to the rooftop setting, low-angle/deep-perspective mood, and includes the technician with transparent umbrella at the frame edge, while preserving the neon-soaked rainy atmosphere. Model B is polished and attractive, but it reads more like a street-level scene than a rooftop and misses the technician presence as a clear subject.)
Drone Parts Burst
A palm-sized inspection drone darting through a narrow server corridor as a maintenance drawer has just been yanked open, sending labels, silica packets, and tiny propeller guards swirling into the air; one rotor is visibly pitching hard, status LEDs streak into short motion trails, dangling ethernet cables arc from the rush of displaced air, and a startled gloved hand reaches in from the right edge; dynamic action photography with high shutter drama and directional cool-white data-center lighting, diagonal composition, 16:9.
![FLUX1.1 [pro]](https://runtimewire.com/api/storage/uploads/showdowns/bb116ae6-8ab4-4337-a818-c60b03fd7c6c.jpg)

Model B adheres more closely to the prompt by clearly showing the maintenance drawer yanked open, swirling labels and packets, dangling ethernet cables, and a gloved hand entering from the right within a believable server-corridor action scene. Model A is more cinematic and polished, but it misses several key prompt specifics and feels less grounded in the described maintenance-drawer moment. (Second judge pass, order swapped — scores are the average of both: Model B adheres more closely to the prompt by clearly showing the maintenance drawer yanked open, dangling ethernet cables arcing through the air, scattered labels, and a gloved hand entering from the right in a believable server-corridor scene. Model A is more immediately dramatic and cinematic, but it misses key prompt specifics like the open drawer contents, tiny propeller guards, and cable motion, and its energy-burst look feels less grounded in the described action photography.)
Matchup powered by OpenRouter.