File: e521ae0c-de26-4a13-bee6-9(...).jpg (3.29 MB, 1792x2400)

3.29 MB JPG

Thanks for the quick bake, anonie.
>>109057457
I quite openly dislike fascism? Mind you, I'm not against prompting swastikas in other settings, for real, I think that letting nazifascists own the symbol is a symbolic defeat, several cultures used that symbol before. But I'm not especially interested in prompting it.

File: download - 2026-06-14T175(...).jpg (83 KB, 768x512)

83 KB JPG

>*laughing in listening to a fascist go on about being anti-fascism*
literally the first reply is that left over from prior thread.. lol, lmao.

File: 9f5f65f4-7960-4d7d-b226-b(...).jpg (2.12 MB, 1792x2400)

2.12 MB JPG

>>109057541
You're the fascist here, anon. You refuse to engage in discourse. And you only attack others. That's the very definition of fascism.

Damn, did goyimAI just cap image generation by a lot? My plus trial ran out of generations until 9PM, a free account ran out after one image

If DALL-3 is dead then how are you guys making these images?

File: _efe300d5-c114-4ef2-bd63-(...).jpg (280 KB, 1792x1024)

280 KB JPG

>>109058197
The thread has been, for quite some time, *also* about "Cloud AI Image Generation", it's in the title. Maybe the slashed handle will have to get updated at some point. Anyway, right now, dalle-3 is still not dead on BIC, at least not on my end. This is a new gen.

File: quick snack.jpg (3.46 MB, 2400x1792)

3.46 MB JPG

>>109048079

File: GPT image 1.png (3.06 MB, 1122x1402)

3.06 MB PNG

Sumi-e inspired ink illustration — a traditional Japanese style emphasizing minimalism, boldness, and spiritual depth through the use of black ink and subtle tonal gradients.
Additional Elements:
Spot Color: A single muted, earthy tone (e.g., deep moss green or burnt umber) used sparingly to highlight key elements — the hunter’s jacket lining, a glint in the scope, or a shadowed accent. Muted Color Palette: Beyond the spot color, the piece remains largely monochromatic, relying on varying ink densities to create depth and mood.

Subject: The Hunter Character: A Caucasian man, mid-30s to early 40s, clean-shaven or with a short, neatly trimmed beard — conveying discipline and focus.

Expression: Serious and determined. His gaze is locked down the scope of the rifle, brows slightly furrowed, jaw set. The ink lines around his eyes are slightly heavier to emphasize intensity.

Attire: Modern Hunting Gear: A high-visibility safety orange collar peeking from under a muted, earth-toned jacket (perhaps charcoal or dark olive), designed for stealth and weather resistance. Layering: A lightweight base layer in soft gray tones, visible at the cuffs and collar, suggesting practicality and experience. High Boots: Black rubber boots with mud splatters rendered in the spot color — grounded, ready for terrain. Gloves: Fingerless, dark leather, functional and worn, showing subtle ink smudges from handling gear. Rifle: Remington Model 700 — rendered with precise, clean lines in deep matte black ink. The stock is walnut, shaded with ink gradients to simulate wood grain and depth. The scope is metallic, reflecting a sliver of the forest in its lens — the only place where the spot color might subtly glow, mimicking a distant glint of light. Setting: The Mysterious Forest Composition: Extreme close-up, focusing on the hunter’s face and the rifle’s scope, with deep foreground and obscured background.

>>109059716
Background:

Trees: Silhouetted in thick, bold Sumi-e strokes — vertical lines suggesting trunks, layered horizontal brushwork for branches. No detail, only shape and suggestion.
Fog/Mist: Rendered with soft ink smudges and graded washes — a hazy, indeterminate depth that pulls the eye into the unknown. Light: A faint, diffused glow filters through the canopy — represented by subtle gradations of ink density, with perhaps a whisper of the spot color in a single shaft of light, echoing the glint in the scope.
Atmosphere: Mysterious, tense, and meditative. The forest is not just a place — it’s a presence. The ink work evokes silence, patience, and the weight of decision.

>>109059722
Sumi-e Techniques Employed
Bokuyō (墨痣): Small ink blots or stains used to suggest texture — on bark, mud, or fabric folds.
Haboku (破墨): “Broken ink” — soft, blurred edges for mist, shadows, and emotional weight.
Mitsuro Bokushi: Layered ink washes to create depth in shadows and distance.
Monolinear Elements: Clean, sharp lines define the rifle, face, and boots — contrast against the softer forest.
Spot Color Application
Purpose: To draw focus and ground the scene in reality.
Placement Ideas: The inner lining of the jacket — a flash of muted green against the black ink.
A drop of condensation on the scope lens. Mud on the boot toe — a single damp patch in burnt umber. A hint of light in the hunter’s eye — a subtle warm tone to give life to the gaze.
Mood & Narrative
Emotion: Stillness before action. The man is not rushing — he is waiting, calculating, respecting the moment.
Theme: The intersection of humanity and nature, precision and patience, the ethical weight of the hunt.
Silent Drama: No sound but the crunch of leaves, no movement but the breath of wind — all conveyed through ink and stillness. Presentation Suggestions
Paper: Cold-pressed rough watercolor paper, to accept ink grain and texture.
Framing: Simple black wood frame — unobtrusive, to keep focus on the artwork.
Lighting: Soft, directional sidelight to mimic the forest glow and emphasize ink shadows.

File: GPT image 2.png (3 MB, 1122x1402)

3 MB PNG

A lone samurai clad in battle-worn armor of dark indigo and rust-red lacquered plates rides bareback on a powerful steed galloping across a windswept moor. The horse is captured mid-leap, muscles taut, mane and tail whipping dramatically in the wind—hooves churning up clods of earth in dynamic motion. The samurai leans forward, one hand gripping the reins, the other raised high, brandishing a gleaming katana that catches the last fiery light of day.

Behind him, the horizon is split by a vivid red setting sun, half-submerged beneath rolling hills. Its rays bleed across the sky in radiant streaks of crimson, burnt orange, and deep ochre, casting long, dramatic shadows that stretch like fingers across the land. Wisps of charcoal-gray clouds drift overhead, adding to the brooding, mysterious atmosphere.

The color palette is rich and grounded: earth tones dominate—umber soil, mossy greens in distant foliage, and weathered browns in the samurai’s armor—yet contrasted with vivid, saturated hues: the glowing sun, the deep vermillion of his under-robe (naga-bakama), and subtle gold leaf accents on sword fittings, catching the dying light.

In true Yoshitoshi style, the composition balances intense action with poetic stillness. Fine woodgrain textures are visible in the background sky and ground, mimicking traditional bokashi (gradient ink) techniques. Delicate linework defines the folds of fabric and the tension in the horse’s form, while slight supernatural elements linger—a faint shadowy figure perhaps in the distance, or the suggestion of ancestral spirits swirling in the wind, typical of Yoshitoshi’s dramatic and often mystical themes. Mood: Heroic, melancholic, and charged with destiny—like a moment frozen between life and legend.

File: GPT image 3.png (3.19 MB, 1086x1448)

3.19 MB PNG

Linocut print, high contrast, bold lines. A meticulously detailed, fortified castle perched upon a jagged, rocky cliff overlooking a dynamic seascape. The scene is dramatic: The sun is positioned directly behind the castle, creating a powerful silhouette and a strong halo effect. Below the cliff, a pristine beach features stylized palm trees (addressing "palmers"). Dense, volumetric clouds fill the sky. The texture must emphasize the characteristic carving marks and heavy ink saturation of a traditional linocut, utilizing dramatic chiaroscuro and stark black-and-white elements. Graphic novel aesthetic, Gustave Doré detail filtered through the medium of Frans Masereel.

Keywords Breakdown:
Component Keywords Used
Style/Medium Linocut print, bold lines, high contrast, black and white, carved texture, graphic aesthetic. Subject Meticulously detailed castle, fortified, jagged rocky cliff, beach, palm trees.
Atmosphere/Lighting Powerful silhouette, sun behind castle, halo effect, dramatic chiaroscuro, dense clouds. Detailing Highly detailed, stylized (for the palms), volumetric.

File: GPT image 4.png (2.64 MB, 1383x1137)

2.64 MB PNG

Supermassive Retrofuturistic Roman Suspension Bridge Vision Description
The Towers
Imagine colossal bridge towers rising 400 meters into the dawn sky, each one a triumphant reimagining of Roman architecture at an impossible scale. The towers are constructed from massive blocks of polished white travertine, so highly reflective they seem to glow with inner light as the morning sun strikes their eastern faces.

Each tower takes the form of a quadruple triumphal arch — four grand arched openings arranged in a cross pattern, allowing the roadways and rail lines to pass through ceremonially. The edges are trimmed with bands of burnished gold that catch fire in the early light, creating brilliant lines that define every architectural element.

Crowning each tower: deep wine-red rooftops in the style of Roman temples, their terracotta tiles darkened to the color of aged Bordeaux, creating a striking contrast against the pale stone and golden accents.

The Statues
Standing sentinel at each tower are eight massive marble figures, each 60 meters tall:
Mars in full military regalia, golden spear raised Neptune commanding the waters below, trident gleaming with gold inlay
Roma herself, personification of the Empire, crowned in gilded laurels Victory with wings spread wide, gold-leafed feathers catching the light The white Carrara marble contrasts magnificently with their golden armor, weapons, and ornamentation.

The Dual Decks
Upper Level: Six lanes of roadway for vehicles, flanked by colonnaded walkways with Corinthian columns

Lower Level: Four rail lines suspended beneath, enclosed in an arched gallery with regular window openings

The morning light transforms everything — the travertine blazes white-gold, the statues cast long shadows westward, and the wine-dark rooftops seem to float above it all like imperial crowns.

File: GPT image 5.png (2.12 MB, 1536x1024)

2.12 MB PNG

70's 16mm grainy photography, 1600 ISO, warm colors, chiaroscuro. An 30 y.o. sicilian don (black hair, closed face and look at you, wearing a black three piece suit, white shirt, tie and signet ring on the ring finger) sitting in the armchair in his luxury office at night, crossed leg.

File: GPT image 6.png (2.16 MB, 1536x1024)

2.16 MB PNG

Minimalist lomography-style painting, intentionally undetailed with soft brush strokes, limited warm color palette (peach, amber, terracotta, soft gold), close-up composition. A comely young anime-style woman with pale skin, long strawberry-blond hair flowing downward, bright blue eyes, and a gentle smile. Slender cyborg arms with subtle metallic sheen, wearing a fitted bodycon gym outfit. She performs a graceful handstand in a dynamic action pose, hair cascading like a sine wave. Setting: a sunlit public garden at midday, soft warm light, blurred floral background, lomography vignette, grainy film texture.

File: GPT image 7.png (3.01 MB, 1536x1024)

3.01 MB PNG

Intricate 1970s anime-style mecha cockpit interior, highly detailed pencil sketch with sharp ink lines and flat pastel coloring. A beautiful young Caucasian woman with pale skin and long chestnut hair, wearing a tactical jumpsuit with sage green and mustard yellow armor, piloting the mecha. Surrounded by analog control panels with dials, switches, and radar screens in soft green/yellow tones. Retro-futuristic mechanical details: exposed pipes, hydraulic lines, and a suspended pilot seat. Vintage cel-shaded aesthetic, cross-hatching textures, cinematic low-angle shot, ultra-detailed, in the style of Go Nagai and classic mecha anime.

File: 82b6ba46-d5b1-4f5a-95b3-1(...).jpg (3.2 MB, 1792x2400)

3.2 MB JPG

>>109059716
>Remington Model 700
Not sure about gpt-image-2, but modern models can probably handle a variety of models. That one was useful in de3 as it looked like one way to trigger something that didn't look like an absolute, nonsensical, abomination with some consistency... Kept it here anyway, your prompt in NBP, just exchanged it with a woman with the usual Dutch braids.

File: 558ad0dd-bd45-4ef6-9c45-f(...).jpg (3.71 MB, 2752x1536)

3.71 MB JPG

Benelli Lupo, 2020 vintage. Almost anachronistic in 1990s style.

Huh, I didn't know bing recycled created images
I copied a prompt I was using on one account and used it on a different account, and suddenly it was creating instantly. But then I noticed they were the same images from earlier

File: IshaMyWife.jpg (1.96 MB, 1920x1080)

1.96 MB JPG

How do I upscale without leaving an oily mess?

File: how to carry megumi kato.png (843 KB, 1200x896)

843 KB PNG

Pretty hard wrangle due to unfamiliarity with pose and filters not liking it. Had to generate with tights first then edit to thigh-highs.

File: FrogRelax.jpg (1.99 MB, 3840x2160)

1.99 MB JPG

I'm losing too much image quality somewhere.

>>109060402
>>109060407
I'd like to try it with an M14 or SKS.

>>109062597
>>109062610
I'm afraid the botched rhinoplasty and Botox is indeed permanent, miss

>>109060723
I find this gen interesting, perhaps because of the anachronistic aspect of the situation.

File: d2799d15-6f56-4453-84b6-f(...).jpg (3.55 MB, 2752x1536)

3.55 MB JPG

>>109063828
>SKS
Asked to remove the bayonet, apparently it's not legal in Russia, and obviously it fails completely (it doesn't seem to understand that that thing attached below and turned inside is the bayonet).

File: a40b54b8-15f3-4b09-97df-1(...).jpg (3.25 MB, 2752x1536)

3.25 MB JPG

Perspective remains another chimera.

File: 05fe8dbe-3e31-4eb5-913a-f(...).jpg (2.95 MB, 2752x1536)

2.95 MB JPG

>M14

File: 1e4d45cc-adef-4b8c-b20c-1(...).jpg (3.1 MB, 2752x1536)

3.1 MB JPG

Here the issue is the magazine, while the prompt asks for the 20 rounds one, sometimes it will generate it like this. NBP likes to do its own things...

>>109064211
>illegal to have a blade attached to a gun
wtf gay
oh, I know, in a land where people are too poor to buy ammo, a blade increases one's capability to continue the fight!

File: kwabl.jpg (107 KB, 792x748)

107 KB JPG

>>109064396
>>109064386
dog shit nonsense as usual

File: 80b3f1a7-0c9b-4576-997c-f(...).jpg (3.38 MB, 2400x1792)

3.38 MB JPG

>>109064645
Russia has some idiosyncratic hunting laws that got recently updated. The most peculiar thing is probably that first timers are restricted to break action shotguns, then after 2 years without incidents they can move to magazine-fed shotguns and only after 5 years they can get a rifle. Outside the former USSR, I think only Japan has something similar (with an even longer, 10 years, shotgun-only probation period).

Talking about shotguns
>687 Silver Pigeon V

File: Krampus001.jpg (2.01 MB, 3870x2160)

2.01 MB JPG

So correct me if I'm wrong. If you are using Gemini, internally it stores data as jpg, so if you make edit after edit to an image, it is re-compressing the unchanged parts every time? Why wouldn't they store data as png?

File: 28fd9f4c-d309-40a8-a3ad-6(...).jpg (3.54 MB, 2400x1792)

3.54 MB JPG

>>109062782
Didn't experiment a lot with it, but that pose seems to get rejected quite strongly even with pants, which is a bit counterintuitive (nonsense about minors and harm).

>>109058197
it's still dead, novelai killed it for good and is a good thing.

>>109065580
princess carry is fine but pic on right is very much a sex position

>>109065905
slop trend
>off shoulder top that isn't anything new and is obviously designed that way

File: 27535f9d-f3e7-4d9c-81d0-1(...).png (2.32 MB, 1536x1024)

2.32 MB PNG

>>109067065
gpt image 2 doesn't quite know how to make Bajorans.

File: 153790cc-9e93-49ae-a250-4(...).jpg (3.2 MB, 2400x1792)

3.2 MB JPG

>>109066441
With two women and no strap-on, it would be more like a parody of a sex position.

Looks like dall-e 3 is now dead on bing. They are doing full prompt rewriting with very long prompts, and the images look like the exact kind of "dall-e 3 substitute" they claimed they were working on. No more mention of dall-e 3, either.

File: screenshot-1781611348084.png (406 KB, 1892x702)

406 KB PNG

>>109068423
It still does mention dalle-3 on my end. And, if the alternative text thing still works as confirmation, it doesn't seem to rewrite long prompts. I suppose it's being rolled out in phases, some people were saying that de3 stopped working on their end already some weeks ago.

>some weeks
Like, two weeks ago, IIRC.

>>109068601
If you're lucky, you still have it. If not, it's already gone.

File: 1e65cfb2-b6d2-4473-ac86-b(...).jpg (3.86 MB, 1792x2400)

3.86 MB JPG

>>109068750
I think roll out of new "features" (if one can call that the removal of a model) in BIC has always been slower here, for better or worse. I wonder what exactly is that replacement, I was half-sure that they were just going to push the idea that MAI-Image-2.5 is good enough to be called a replacement, not that they were going for aesthetics genuinely similar to de3.

Talking about removal of old models, Imagen 4 is effectively gone from Flow on my end.

File: 535e8686-e6b4-4f18-8478-0(...).jpg (2.83 MB, 2752x1536)

2.83 MB JPG

Actually, it looks Imagen 4 was missing only from one account (or I was seeing things/interface got cut out somehow). My bad.

On another note, NBP tends to make void car plates or it fills them with obvious placeholders, unless one specifically requires a realistic card plate, but it tends to reproduce quite often the ones from reference photos, so much for privacy (I had to reference this configuration, it's rare enough that NBP can't generate it directly).

>>109068972
I discovered it still works on mobile, someone mentioned it on /r/bing. So it works on my phone. For now, anyway.

File: Soft_Pre-Raphaelite_paint(...).jpg (3.36 MB, 3520x2336)

3.36 MB JPG

>be me
>prompt for four cloud leopards
>get three snow leopards
Life's a beach and then you die.

>>109067726
>>109068437
Nice skyscraper in center of gens.
The second one reminds me of the pyramidal shape of Tyrell Corp's headquarters in Blade Runner, only smaller, without the buildings surrounding the main part, and without the upper part inspired by ancient Rome.

What were your inspirations?

File: e754d3e7-c091-4c7a-9550-5(...).jpg (3.85 MB, 2400x1792)

3.85 MB JPG

Imagen 4 seems to be gone for real now and, curiously enough, the percentage indicator has returned (with NBP gens being completed at little more than 50% again).

>>109068423
I'm seeing it too. Benchmark prompts are producing substantially different results. More burnt, more flat, generic poses, less sensitive to requested art styles.

Not happy.

>>109071971
also, getting BAD sameface issues

I generated some lewd images with meta AI and they brutally nerfed my account lol. They removed the fork feature where one can return and edit previous parts of the conversation. Also now anything remotely suggestive but still harmless looking can't be generated. And strangest of all it's like I was switched to a deprecated model that constantly gets things wrong, does lower resolution images with weird facial distortion effects. Really odd way of avoiding just suspending someone's account.

>>109071980
It's a trash finetune of MAI-image-2, as expected. RIP Dall-e 3, for reals this time.

File: its-ogre.png (25 KB, 430x322)

25 KB PNG

>>109072368
yep it's ogre

File: Snail_girl_is_content._Ba(...).jpg (2.68 MB, 2752x1536)

2.68 MB JPG

Exploring the interface of google flow. Not completely disliking it, even if it refuses to accurately follow model sheets when making videos.

how many free images on flow now?

File: _2d40fade-835d-4ba3-949d-(...).jpg (268 KB, 1024x1536)

268 KB JPG

>>109073270
I thought the "limit" was only with Pro. Even with Nano Banana 2 I sometimes get hit with "You are doing that too much" that can last for hours, but I assume that might be a different thing

File: They_are_playing_monopoly(...).jpg (2.8 MB, 2752x1536)

2.8 MB JPG

Interesting how in flow you can summon random characters quickly once you have the models inputted. A pity that the quality of the output gets compressed looking, super extremely quickly, if you make any edits at all after the initial output.