
Flux.2 Klein 4B
Best-in-class instruction following with a strong balance of quality and responsiveness
- Image Quality
- Generation Speed
- System Requirements
- Medium
- Text Generation
- No


PromptA solitary astronaut wearing a sleek, metallic black space suit stands on a surreal, alien world bathed in the golden hues of dawn, framed by a vast, vibrant landscape.


PromptCyberpunk Medusa robot female, seductive gaze, shiny jewelry glittering under dramatic light, vibrant colors, hyper-realistic Octane render style.


PromptA mystical wizard stands in the foreground, his weathered face is a blend of intensity and mystery, with sharp features illuminated by soft, golden light spilling from above.
Choose what’s right for you, and your PC.
Not all models produce the same results. Not all PCs have the same performance, speed and memory access. Distinct AI allows you to choose the models that will give you the local results you’re looking for.

Best-in-class instruction following with a strong balance of quality and responsiveness

Top-tier visual quality and text rendering, prioritizing image fidelity at reasonable speed

Polished visuals suitable for portraits, creative art and concept work.

Ultra-fast image generation

A great all-rounder for everyday image generation, with solid speed and results

Reliable, well-balanced image quality with a neutral and realistic style

Primarily for artistic expression rather than photo-realism, with visually rich results

Focuses on higher realism and improved visual fidelity, with quality results

Premium-quality visuals with faster generation speed than the Large Model

Maximum image quality, exceptional detail, composition and realism

Purpose-built for anime and manga styles producing vibrant character designs

High-quality, modern visuals with strong text rendering capabilities

High visual quality and decent text rendering, versatile and refined

Industry-leading image quality, realism, prompt understanding and typography

High visual quality and decent text rendering, versatile and refined

Purpose-built for image editing and transformation workflows

Exceptional instruction following and flagship image-editing model
Unlimited AI image generation. No credits required.











PromptA majestic lion resting on a rocky outcrop at sunrise, golden-hour lighting, vast hills blurred in background...

PromptA florist with wavy auburn hair arranging flowers, peonies, wildflowers, in a florist shop, realism...

PromptA ballerina dancing, posing on a large stage, charcoal drawing style, black-and-white, light coming in above...

PromptA city of white-domed rooftop buidlings, growing among enormous wildflowers, illustrative, colorful, cheerful...
Flux.2 Klein 4B combines premium image quality with advanced instruction-based workflows. It excels at image editing, image transformation, prompt-driven refinement, and creative control. Its ability to understand complex requests and make meaningful visual changes makes it one of the most versatile models in the Model Hub and a favorite for both image generation and image editing tasks.





PromptCloseup shot of a hopeful robot warrior with rain-soaked fabric in the rain, on a muddy road, cityscape...

PromptA wide-angle perspective, side-facing lighthouse keeper wearing a yellow raincoat during a storm, lighthouse...

PromptA punk rock fan, teenage girl, a break from studying in her room, posing, laptop open and posters on the wall...

PromptA strikingly elegant Mongolian princess, luminous skin, ornate gold armor dress, low-angle shot, lake at night...
Z-Image Turbo delivers flagship-quality image generation while maintaining practical generation speeds. It produces highly realistic imagery, excellent prompt adherence, and strong text generation capabilities. For most users, Z-Image Turbo is the recommended version of Z-Image thanks to its superior balance of quality, realism, and usability.



PromptPortrait of a model, night scene, wearing a liquid porcelain sapphire malachite ruby mask, deep green...

PromptA man in a scenic environment by Martin Schoeller, cinematic lighting, focused, natural background...

PromptPurple crow holds a pink lollipop, surreal fantasy style, soft spotlight illumination, macro, closeup...

PromptJoyful young woman with straight, black bob haircut, selfie, pink outfit in front of a merry-go-round...
DreamShaper XL Lightning excels at stylized and polished visuals while maintaining impressive generation speed. It produces vibrant illustrations, fantasy environments, character designs, and creative concept artwork with a distinctive artistic flair. For creators seeking imaginative visuals rather than strict realism, DreamShaper XL Lightning is a standout choice.







PromptA futuristic robot walking towards us in the middle of the street in a major city...

PromptCloseup portrait of a muscular, bearded man in a worn steel mech suit, rusted with sharp focus...

PromptPowerful lion in vector art, vibrant yellows, greens, blues, high contrast shapes, bold geometric lines...

PromptWhimsical fairy with big eyes, long eyelashes, green dress, focused on face, watercolor style...
SDXL-Turbo is designed for maximum speed, generating images in seconds with minimal hardware requirements. It excels at rapid concept development, previews, storyboarding, and batch generation where speed matters more than ultimate visual fidelity. If you want to explore ideas quickly, SDXL-Turbo is one of the best starting points available.







PromptFront view of puppy, realistic photography, even soft lighting, textured fur, meticulous detail...

PromptColossal robot vehicle, cyberpunk style, low-angle, neon lighting, fururistic city, digital painting...

PromptA youthful woman with curly hair and freckles stands in a sun-dappled garden, golden light filtering in...

PromptBeautiful alien woman with multicolored hair, surreal geometric prism leaves, kaleidoscopic color palette...
SSD-1B offers an excellent balance between image quality and generation speed, making it one of the most versatile models in the Model Hub. It produces noticeably better results than SDXL-Turbo while maintaining responsive performance and broad hardware compatibility. For most users, SSD-1B is the ideal everyday image generation model.







PromptCinematic scene: Giant scaly lizard monster, sharp teeth, foggy, attacking downtown city, foggy atmosphere...

PromptBeautiful alien woman with multicolored hair, surreal geometric prism leaves, light flairs, Mark Dion style...

PromptCloseup portrait of young boy with freckles, natural smile, soft morning light, realistic oil painting...

PromptParrot amidst a hidden sanctuary garden, fantasy art concept, whimsical, detailed closeup...
SDXL 1.0 Base delivers reliable image quality with a neutral, realistic style. Often considered the Swiss Army Knife of image generation models, it performs well across a wide variety of subjects and artistic styles. Its consistency and versatility make it a dependable option for creators who want one model that can handle almost anything.







PromptBeautiful alien woman with multicolored hair, surreal geometric prism leaves, glowing mushrooms...

PromptVegetarian pizza, on a gingham tablecloth, Italian restaurant, window light, realistic style...

PromptAnime-style teenage hero, dynamic flowing silver hair, Studio Ghibi inspired, closeup, vibrant neon accents...

PromptAnime style, young summoner casting spell, glowing magic portal in secluded swamp, dramatic lighting...
Playground 2.5 emphasizes creativity, artistic direction, and visual storytelling. Rather than focusing purely on realism, it produces imaginative and visually rich imagery with strong composition and artistic flair. It is particularly well suited for creators who value style, atmosphere, and creative expression.







PromptRealistic calico cat with pale eyes wearing a knit scarf standing on a stone street in London...

PromptAggressive red American muscle car, sleek and powerful, under neon lights, foggy, raining, realistic...

PromptHyper-realistic steampunk woman wearing intricate brass and leather hat, moody atmosphere...

PromptA towering robotic figure with detailed metallic armor strides through a neon-lit cityscape at night...
Stable Diffusion 3.5 Medium delivers a significant leap in realism, prompt adherence, and overall image quality. It excels at producing polished, professional-looking visuals while remaining accessible to a broad range of systems. For creators looking to move beyond SDXL-era models, SD 3.5 Medium offers a substantial upgrade in image fidelity and prompt responsiveness.
* Requires an NVIDIA RTX 30-series GPU or newer.







PromptA floating observatory suspended in a stormy, cloudy sky with lightning in the background and lighting below...

PromptA fashion model in a red coat, at night, walking on the street in Tokyo, wet streets, bright lights...

PromptA rugged fisherman repairs nets in a fishing harbor with boats in the water in the background...

PromptA closeup of a playful red panda climbing through bamboo trees in the rainforest...
Stable Diffusion 3.5 Large Turbo combines premium image quality with significantly improved speed compared to the full Large model. It produces detailed, polished visuals while maintaining strong prompt understanding and text capabilities. For many users, this represents the sweet spot between performance and image quality within the Stable Diffusion family.







PromptA futuristic female pilot standing in a hangar beneath a colossal spaceship, realism, detailed...

PromptA monarch butterfly resting on a vivid sunflower, late afternoon, closeup, macro, hyper-realistic photo...

PromptA desert nomad riding across the desert on a mechanical robot camel, alien planet, two moons...

PromptA traveler woman watching hot air balloons, magic hour, photographic quality, realism, golden sunset...
Stable Diffusion 3.5 Large delivers flagship visual quality with exceptional detail, composition, and prompt understanding. It is designed for users who want the highest possible image fidelity and are willing to trade generation speed for superior output. In most cases, Large Turbo will be the better overall choice, but SD 3.5 Large remains a powerful option when maximum quality is the priority.







PromptGirl solo, cyberpunk hacker, night, neon signs, manga style, vivid colors, costume...

PromptSolo girl fairy with tiny wings, with a flower dress, sitting on a mushroom-like seat, deep in the forest...

PromptMechanical robot battle, clashing in a ruined city with explosions and action in the background...

PromptBento lunch box, rice balls, tamagoyaki, vegetables, colorful blanket, picnic on a sunny day, closeup...
Animagine XL 4.0 is specifically designed for anime-inspired artwork, manga illustrations, fantasy characters, and stylized storytelling. It produces vibrant character designs, expressive faces, and polished anime aesthetics while maintaining excellent consistency. For anime creators and fans, this is one of the strongest specialized models available.







PromptA solitary Korean woman standing in a moody sumi-e vignette, her slender body framed by a dark stormy sky...

PromptA male surfer in his late 20s stands on the beach, golden hour lighting, surfboard craddled under his arm...

PromptAn enormous floating fantasy city hovers above a swirling storm cloud, iron and glass towers, domes...

PromptA young woman with curly chestnut-brown hair, freckles, stands in a meadow by a fence, raising her arms...
Flux.1 Schnell is one of the strongest realism-focused models available in the Model Hub. It consistently produces highly detailed imagery with natural lighting, strong composition, and impressive prompt adherence. While capable of artistic output, its greatest strength lies in creating photorealistic scenes and commercial-quality visuals.







PromptA young woman in a meadow with curly, sun-kissed hair cascading in soft waves, natural tan, freckles...

PromptA hyper-realistic image of a closeup of a pig on soft, ethereal pink clouds, low-camera angle, detailed...

PromptA photo-realistic portrait of a strange blue alien creature, captured in a low-angle shot, shallow depth of field...

PromptA black woman captured with a 50mm lens at f1.8 aperture, closeup, rich burgundy hair and lip gloss...
Z-Image focuses on producing highly polished visuals with exceptional detail, image fidelity, and strong prompt understanding. It delivers excellent text generation capabilities while maintaining flexibility for experimentation and creative refinement. Compared to Z-Image Turbo, it prioritizes control and image quality over generation speed.



PromptA skeletal creature with a brass multi-lens camera head and trailing cables, wading through a flooded church...

PromptA night forest of giant glowing mushrooms, drifting spores, and a young deer beside a still black pond...

PromptAn enormous humpback whale drifting through the clouds above a mountain village at golden hour...

PromptInside a glacier ice cave of glowing blue walls, a shaft of daylight, and a tiny climber...
Qwen Image 2512 Lightning delivers industry-leading image quality, realism, prompt understanding, and typography capabilities. It maintains most of the visual quality of the full Qwen model while dramatically improving usability through faster generation speeds. For many creators, this represents the best overall model in the entire Model Hub.
* Requires an NVIDIA RTX 30-series GPU or newer.



PromptA huge Bengal tiger walking toward the camera through knee-deep water in a flooded jungle temple at dawn...

PromptAn abandoned church flooded with water, stained-glass light through haze, trees growing through the pews...

PromptA gigantic great white shark exploding out of a stormy sea, jaws wide open, water streaming off...

PromptA lone traveller crossing a rope bridge over a jungle canyon, with a huge waterfall thundering beside him...
Qwen Image 2512 represents the pinnacle of image quality within the Model Hub. It produces exceptional realism, outstanding prompt adherence, industry-leading text generation, and highly refined visual output. While it requires powerful hardware and sacrifices generation speed, it delivers the highest-quality results available.
* Requires an NVIDIA RTX 30-series GPU or newer.



PromptAdd round yellow sunglasses and change her hair to jet black, keeping the same hairstyle and everything else...

PromptAdd a small orange tabby cat sitting on the wooden bench on the right, keeping everything else unchanged...

PromptChange the scene from night to a bright winter morning with a clear blue sky and sunlit snow...

PromptReplace the bamboo forest with a snowy mountain temple courtyard, keeping the samurai and pose unchanged...
Qwen Image Edit 2511 Lightning is purpose-built for image editing and transformation workflows. Rather than generating images from scratch, it excels at understanding editing instructions, preserving image quality, and making intelligent modifications to existing visuals. The Lightning version delivers faster performance while maintaining most of the quality of the full model.
* Requires an NVIDIA RTX 30-series GPU or newer.



PromptChange only the sky: replace the sunrise with heavy dark storm clouds, keeping the man and pier unchanged...

PromptAdd a big red polka-dot party hat with a chin strap, and three lit cupcakes on the rug...

PromptChange only the orange accents on the car to electric blue, keeping the black body, garage and lighting...

PromptRemove the astronaut and make the greenhouse look abandoned, with thick vines and cracked glass...
Qwen Image Edit 2511 is the highest-quality image editing model in the Model Hub. Designed specifically for image transformation and enhancement, it delivers exceptional prompt adherence, visual consistency, text generation, and image fidelity. It prioritizes maximum editing quality over speed, making it ideal for demanding professional workflows.
* Requires an NVIDIA RTX 30-series GPU or newer.


Select at least one model with the Compare checkbox on a card, then reopen this.
* Requires an NVIDIA RTX 30-series GPU or newer (BF16). All other models require RTX 20-series or newer.