Gemini 3 Pro with image generation capabilities. Combines advanced reasoning with the ability to generate and edit images.
Settled by community votes across 19 shared challenges, with an AI judge weighing in on each.
Nano Banana Pro
#2 of 62 in Text-to-Image
Wan 2.5 (Preview)
#27 of 62 in Text-to-Image
Where the votes landed
Nano Banana Pro
0%
win rate
Ties
0%
Wan 2.5 (Preview)
0%
win rate
Challenge by challenge
The strongest take from each model on every shared challenge, with the AI judge's read.
Geometric Composition
Text-to-Image“A glass cube on a wooden table. Inside the cube is a small blue sphere. On top of the cube sits a red book. A green plant is behind the cube, partially visible through the glass. Soft window light from the left.”
AI Judge Analysis
Nano Banana Pro
- + Excellent photorealism with a film-like quality.
- + Accurate glass physics and reflections on the glass and wooden table.
- + Perfectly captures the lighting direction requested in the prompt.
- − The sphere is quite small relative to the cube size.
- − The book's placement looks slightly floating/unbalanced on the thin glass edges.
Wan 2.5 (Preview)
- + Strong composition and vibrant color palette.
- + Good interpretation of the 'small blue sphere' making it a focal point.
- + Atmospheric dust motes enhance the soft window lighting.
- − The red book is clipping through the top of the glass cube.
- − The perspective of the cube's base feels slightly skewed compared to the table edge.
Verdict: Nano Banana Pro produces a much more realistic image with convincing textures and physics, particularly in how the glass interacts with the environment. Wan 2.5 (Preview) has a pleasing color grade and composition, but fails on physical coherence as the red book appears to merge into the top of the glass cube.
Man and Car in California
Editing“Make a photo of the man driving the car down the California coastline”
AI Judge Analysis
Nano Banana Pro
- + Excellent preservation of the man's facial features and specific hairstyle.
- + Faithfully replicates the man's plaid coat and black scarf from the source image.
- + Dynamic composition with beautiful golden hour lighting on the California coast.
- − The man is sitting on the right-hand side of the car, which is inconsistent with the driver's side steering wheel position.
- − The car's interior trim details on the door and dash have been modified from the original.
Wan 2.5 (Preview)
- + Preserves the full exterior of the car more accurately in relation to the source image.
- + Correctly places the driver on the left side of the vehicle.
- − The man's face and clothing details are largely lost due to the distance and shadow.
- − The man's distinct hairstyle is significantly simplified compared to the source.
Verdict: Nano Banana Pro successfully combines the subjects of both source images with high fidelity, maintaining the specific clothing and facial features of the man, despite a confusing seating orientation. Wan 2.5 (Preview) provides a more realistic driving scene in terms of scale and positioning, but it fails to preserve the unique identity and details of the person requested.
Candid Street Photography
Text-to-Image“A candid street photo of an elderly Japanese man repairing a red bicycle in light rain, reflections on wet pavement, shallow depth of field, 50mm lens, natural skin texture, imperfect framing, motion blur from passing cars, cinematic but realistic, no stylization.”
AI Judge Analysis
Nano Banana Pro
- + Captures the 'candid' and 'imperfect framing' request perfectly with a realistic street photography aesthetic.
- + Excellent city atmosphere with authentic Japanese street signs and realistic wet pavement reflections.
- + Features realistic motion blur from passing cars as specifically requested.
- − The man's right hand has some anatomical merging issues with the bicycle components.
- − The bicycle's rear rack and seat assembly are structurally slightly nonsensical.
Wan 2.5 (Preview)
- + High resolution and very clean, sharp focus on the subject's face and hands.
- + Convincing rain droplets on the red paint of the bicycle.
- + Better rendering of small tools on the ground which adds to the narrative.
- − Fails to include the requested 'motion blur from passing cars', as the background vehicle is stationary or frozen.
- − The lighting and skin look slightly more 'digitally clean' despite the request for no stylization.
Verdict: Nano Banana Pro is the winner for its superior adherence to the 'candid' and 'imperfect' photography style, accurately capturing the motion blur and busy street atmospheric details requested in the prompt. While Wan 2.5 (Preview) produces a cleaner image, it ignores the specific motion blur requirement and feels more like a staged portrait than a candid street photo.
Fantasy Warrior
Text-to-Image“Close portrait of a battle-worn paladin in ornate engraved plate armor, hair braided with small beads, faint scars and dirt on the skin, warm torchlight reflecting off metal, shallow depth of field, bokeh sparks, lifelike eyes, highly detailed texture on leather straps and cloth underlayer.”
AI Judge Analysis
Nano Banana Pro
- + Excellent adherence to the 'battle-worn' descriptor with realistic grime and skin abrasions.
- + Superior engraving detail on the plate armor that looks naturally integrated.
- + Dynamic lighting with warm torchlight glints on the metal and hair beads.
- − Leather strap texture is slightly softer compared to the sharpness of the metal.
Wan 2.5 (Preview)
- + High-contrast texture on the cloth underlayer and leather straps.
- + Distinct beard and braided hair detail with clear bead placements.
- + Good use of chainmail and depth of field.
- − Skin textures appear slightly smoother and more plastic-like than the gritty look of the other model.
- − The engraving on the armor looks a bit like a flat decal rather than carved into the metal.
Verdict: Nano Banana Pro captures the prompt's essence more effectively, delivering a truly battle-worn appearance with superior skin and armor textures. While Wan 2.5 (Preview) provides excellent detail on the clothing and straps, its skin rendering feels a bit too clean and processed for a paladin in the heat of battle.
Modern Clean Menu
Text-to-Image“Modern minimalist restaurant menu design, white background with colorful food photos in grid, sections for appetizers/pizza/mains, bold sans-serif fonts, vibrant accents, clean professional layout for casual dining.”
AI Judge Analysis
Nano Banana Pro
- + Excellent typography with mostly legible words and clean sans-serif fonts.
- + High-quality, distinct food photography that fills each grid cell effectively.
- + Clear sectioning for Appetizers, Pizza, and Mains as requested.
- − The 'Mains' section contains a pasta dish misleadingly labeled 'Graoy Salad'.
- − Some items lack logical prices or descriptions.
Wan 2.5 (Preview)
- + Attractive minimalist aesthetic with consistent white backgrounds for all food items.
- + Strong use of vibrant accents in the layout and color-coded section dividers.
- + Good variety of food presented in a clean grid.
- − Text rendering is very poor with significant gibberish and artifacts.
- − The pizza section features salads rather than pizzas under most thumbnails.
Verdict: Nano Banana Pro is the winner as it produces a highly functional design with mostly legible text and accurate food representations for each section. While Wan 2.5 (Preview) has a very high-end minimalist appearance, the text is completely illegible and the content of the sections (like putting salads in the pizza category) does not follow the user's navigational requirements.
Magic Burger Explosion: Fiery Photorealism Challenge
Text-to-Image“Ad for 'Magic Burger'. Dynamic, exploded burger with all components (bun, patty, cheese, lettuce, tomato, sauce) suspended in mid-air. Emphasize photorealistic detail and a sense of motion. Dark, fiery background with glowing embers. Integrate text: 'MAGIC BURGER' as a prominent title, 'LIMITED TIME ONLY' as a secondary message, and '€6.99' in a starburst, all rendered with a fiery, glowing effect.”
AI Judge Analysis
Nano Banana Pro
- + Excellent photorealistic textures, particularly on the grilled patty and toasted bun.
- + Flawless text rendering that matches the requested fiery, glowing neon style.
- + Dynamic composition with realistic sauce splashes and crumbs that enhance the sense of motion.
- − The background is slightly more uniform in color compared to the depth in Model B.
Wan 2.5 (Preview)
- + Creative typography with a melting/dripping effect that adds to the 'Magic' theme.
- + Good use of atmospheric smoke and varied debris to fill the frame.
- + Strong color contrast between the cool smoke and warm fire elements.
- − The patty and buns look slightly more plastic and less realistic than in Model A.
- − The price text loses the requested fiery glow, appearing as flat white text inside the starburst.
Verdict: Nano Banana Pro is the winner due to its superior photorealistic textures and perfect adherence to the text styling requirements, creating a high-end commercial look. While Wan 2.5 (Preview) offers a more stylized and artistic font, it fails to maintain the glowing effect on the price tag and the food components look slightly less authentic.
Chalkboard Menu
Text-to-Image“Handwritten-style chalkboard menu in a cozy café, all text rendered in the exact same realistic chalk handwriting style with natural variations in letter size, slight slant, and chalk texture — no printed or digital fonts anywhere on the board. Title at the top in elegant cursive chalk handwriting: ‘TODAY’S SPECIALS – APRIL 30, 2026’. Below it, three menu items also in the same handwritten chalk style: ‘Truffle Mushroom Risotto – $24’, ‘Grilled Octopus with Lemon & Herbs – $28’, ‘Brown Butter Chocolate Chip Cookies – $9’. At the very bottom, smaller text in the identical handwritten chalk style (slightly smaller but still clearly legible with the same handwriting characteristics): ‘All items made fresh daily • Ask about our gluten-free options’. Warm ambient café lighting, visible chalk dust and smudges, realistic handwriting imperfections, no clean printed text anywhere.”
AI Judge Analysis
Nano Banana Pro
- + Excellent text accuracy including the specific date and full phrases.
- + The chalk smudges and texture on the board look very realistic.
- + The composition feels natural within a café setting.
- − The font looks slightly too uniform and digital despite the prompt's request for natural variations.
- − The title is not in 'elegant cursive' as requested.
Wan 2.5 (Preview)
- + The handwriting style feels much more like authentic chalk lettering with varying thickness.
- + The board has a more dynamic perspective and higher contrast.
- − Missed the word 'Herbs' in the second menu item.
- − Repeats the price '$9' twice for the cookies, creating redundancy.
- − Failed to provide an elegant cursive font for the title.
Verdict: Nano Banana Pro followed the text prompt more accurately, correctly including all ingredients and the date without repetition. However, Wan 2.5 (Preview) produced a more stylistically convincing chalk texture and handwriting style, even though it made several errors in the actual text content.
The Reversed Rodeo
Text-to-Image“Horse riding astronaut in space — horse on top, not vice versa. Surreal, highly detailed, cinematic.”
AI Judge Analysis
Nano Banana Pro
- + Follows the specific prompt instruction of having the horse on top of the astronaut.
- + Captures the surreal quality requested with a creative interpretation using a celestial cape.
- + Good background composition with varied celestial elements like galaxies and satellites.
- − The anatomy where the astronaut and horse connect is slightly awkward and lacks physical weight.
- − The horse's front right leg appears slightly distorted where it meets the astronaut.
Wan 2.5 (Preview)
- + Technically high visual quality with sharp lighting and textures.
- + The horse and astronaut models are rendered with great detail and realistic materials.
- − Completely failed the negative constraint/instruction to have the horse on top.
- − Produced a conventional 'astronaut riding a horse' image despite the counter-intuitive prompt.
Verdict: Nano Banana Pro followed the complex prompt instruction perfectly, placing the horse on top of the astronaut to create a truly surreal image. Wan 2.5 (Preview) produced a very high-quality image, but ignored the specific inverse-riding instruction, resulting in a standard interpretation that failed the primary challenge.
Outfit Transfer Challenge
Editing“Use Image 1 as the base person. Dress them in the exact elaborate outfit from Image 2 (including all layers, accessories, jewelry, and shoes). Carefully adapt the clothing to the body shape and pose in Image 1 while maintaining realistic fabric behavior, correct proportions, and perfect lighting/shadow matching. Keep the person’s exact face, hair, and background completely unchanged.”
AI Judge Analysis
Nano Banana Pro
- + Perfectly preserves the person's unique face, hair, and vitiligo patterns.
- + Accurately recreates the pea coat, scarf, and watch from Image 2.
- + Followed all constraints regarding background preservation.
- − The transition between the neck and the coat is slightly rough with a small white artifact.
- − Missed the sunglasses from the second source image.
Wan 2.5 (Preview)
- + Excellent fabric texture and realistic lighting on the coat.
- + Included the sunglasses from Image 2.
- − Failed the most critical instruction by completely changing the person's face and skin.
- − Did not preserve the original person's unique features (vitiligo, hair, and sand texture).
- − The hand holding the watch is anatomically warped.
Verdict: Nano Banana Pro successfully completed the complex task of dressing a specific subject in a new outfit while keeping their physical identity and the background entirely intact. Wan 2.5 (Preview) failed the editing mission by replacing the person in the image with the person from the reference outfit image, ignoring the requirement to keep the base person unchanged.
The Capybara Taxi Driver
Text-to-Image“Photorealistic scene inside a yellow New York taxi at night. A capybara is driving, wearing a yellow taxi driver cap and a dark jacket. It has a calm, professional expression and both front paws on the steering wheel. In the back seat sits a human businesswoman in a coat, looking at her phone with a completely normal, bored expression (as if this is just another normal ride). Through the windows you can see the streets of Manhattan at night with blurred lights. Realistic taxi interior, photorealistic, detailed fur and fabric, 35mm lens, night lighting with reflections, shallow depth of field.”
AI Judge Analysis
Nano Banana Pro
- + Excellent photorealistic lighting and texture, especially on the capybara's fur.
- + Very natural interaction between the paws and the steering wheel.
- + The passenger's bored expression perfectly captures the 'normal ride' instruction.
- − The composition has a slightly cramped feel with the dashboard meter blocking the view.
Wan 2.5 (Preview)
- + Vibrant, cinematic lighting with clear Manhattan background elements.
- + Good adherence to the clothing and cap requirements.
- + Clean, professional-looking dashboard and taxi meter.
- − The paws look somewhat claw-like and less natural on the steering wheel.
- − The scale of the capybara relative to the car seat feels slightly small.
Verdict: Nano Banana Pro produces a more grounded, photorealistic image with superior textures and a better emotional performance from the human passenger. While Wan 2.5 (Preview) offers a cleaner and more colorful composition, Nano Banana Pro's lighting and realistic integration of the capybara into the environment make it the stronger overall choice.
The Halloween Invitation
Text-to-Image“Vintage gothic Halloween party invitation. Dark parchment poster, spooky border with webs and thorns, central glowing jack-o-lantern, bats, twisted trees, moody night sky. Add elegant gothic title text saying "Halloween Party Invitation", a small scroll banner saying "You are invited to a night of frights", and event details at the bottom: Date: 30.10.2026 Time: 7pm Location: The Arches, NYC Spooky but polished, cinematic lighting, square format.”
AI Judge Analysis
Nano Banana Pro
- + Perfect text accuracy for all requested fields including specific dates and slogans.
- + Exquisite vintage parchment texture and moody, atmospheric lighting that fits the gothic theme.
- + The glowing orange text effects tie the typography into the central jack-o-lantern theme beautifully.
- − The border thorns are a bit repetitive and less detailed than the rest of the illustration.
Wan 2.5 (Preview)
- + High visual clarity with sharp details on the pumpkin's internal fire effects.
- + Good use of color contrast between the deep blue night sky and the orange pumpkin.
- + Accurate text rendering for the title and date.
- − Missing the comma in 'The Arches NYC' which was requested in the prompt.
- − The composition feels slightly more digital and less like a 'vintage parchment poster' compared to model A.
- − The thorns on the border appear somewhat synthetic and float over the background.
Verdict: Nano Banana Pro delivered a superior vintage aesthetic with perfect adherence to the text requirements and a more cohesive gothic atmosphere. While Wan 2.5 (Preview) produced a sharp image with impressive fire effects, it lacked the authentic parchment texture and slight punctuation accuracy found in Nano Banana Pro's output.
Bald man challenge
Image Editing“Give the person a full, thick head of natural hair with realistic texture, density, and a natural hairline. Preserve facial features and lighting.”
AI Judge Analysis
Nano Banana Pro
- + Excellent texture and flyaway hair details that match the outdoor environment
- + Integrates the forehead skin and hairline seamlessly
- + Perfectly preserves the facial features and lighting from the source
- − None notable
Wan 2.5 (Preview)
- + Successfully adds a full head of hair that follows the head shape
- + Preserves the original identity and background well
- − The hair texture looks slightly stylized and 'helmet-like' compared to the source beard
- − The hairline is very sharp and less natural than Model A
Verdict: Nano Banana Pro provides a much more realistic result, with messy flyaway hairs and a natural hairline that feels integrated into the environment. Wan 2.5 (Preview) adds the hair effectively, but the texture is too smooth and the hairline is too neat, making it look like a toupee rather than natural growth.
Isometric Miniature Diorama Scenes
Text-to-Image“Create a clear, 45° top-down isometric miniature 3D cartoon scene of Japan's signature dish: sushi, with soft refined textures, realistic PBR materials, gentle lighting, on a small raised diorama base with minimal garnish and plate. Solid light blue background. At top-center: 'JAPAN' in large bold text, 'SUSHI' below it, small flag icon. Perfectly centered, ultra-clean, high-clarity, square format.”
AI judge analysis unavailable for this challenge.
Over-the-top cartoon caricature
Editing“Create a caricature of me and my job. Make it exaggerated and humorous, incorporating my profession as a tv show anchor and my love for dogs and hockey.”
AI Judge Analysis
Nano Banana Pro
- + Excellent caricature style that maintains a strong likeness to the source person.
- + Highly detailed and creative integration of all prompts, including hockey equipment on the desk and dogs in jerseys.
- + Very clean text rendering on the mug and posters.
- − The hockey stick held in the left hand is slightly floating above the palm.
- − A small visual glitch in the 'Breaking News' screen where a hockey stick cuts into the graphic frame.
Wan 2.5 (Preview)
- + Captures the 'exaggerated' and 'humorous' instruction well with wide-eyed expressions.
- + Good integration of the hockey theme via the TV screen and jersey-wearing dog.
- − The likeness to the woman in the source image is less accurate than Model A.
- − The composition feels slightly more generic and less densely packed with the requested details.
- − Anatomical issue with the right hand holding the microphone (odd thumb placement).
Verdict: Nano Banana Pro is the clear winner for its superior ability to maintain the subject's likeness while creating a vibrant, detail-rich caricature. It creatively fills the scene with professional and personal touches, unlike Wan 2.5 (Preview), which felt more like a generic template with less fidelity to the original person's face.
Adorable Baby Animals in Sunny Meadow
Text-to-Image“Hyper-photorealistic scene of fluffy baby animals—a golden retriever puppy, tabby kitten, baby bunny, and red fox kit—with big expressive eyes and ultra-detailed soft fur, playfully chasing butterflies and tumbling together in a lush wildflower meadow, warm golden sunrise light with god rays and dew sparkles, joyful wholesome vibe, 8K masterpiece.”
AI Judge Analysis
Nano Banana Pro
- + Excellent fur texture and detail across all species
- + Cohesive lighting with well-rendered god rays and sparkles
- + Accurate representation of a golden retriever puppy's features
- − The fox looks a bit too similar to the puppy in facial structure
- − The butterfly size is slightly disproportionate to the animals
Wan 2.5 (Preview)
- + Strong sense of movement and 'tumbling' motion in the animals
- + Includes multiple butterflies as requested in the prompt
- + Beautifully rendered dew drops and dandelion seeds adding to the atmosphere
- − The fox's eyes have unnatural, blue glowing artifacts
- − Anatomy issues on the kitten's front right leg which appears disjointed
- − The white rabbit lacks the soft, ultra-detailed fur texture seen in other subjects
Verdict: Nano Banana Pro produces a much more polished and 'hyper-photorealistic' image with superior fur textures and anatomical consistency. While Wan 2.5 (Preview) captures the energetic movement and atmospheric elements like dew better, it suffers from significant artifacts in the fox's eyes and the kitten's limb structure.
Studio Ghibli Anime Style
Editing“Transform this photo into a Studio Ghibli–inspired illustration. Use soft pastel colors, hand-painted textures, gentle lighting, dreamy backgrounds, and a warm, nostalgic mood”
AI Judge Analysis
Nano Banana Pro
- + Excellent hand-painted watercolor texture that fits the prompt perfectly
- + Successfully transforms the background into a dreamy rural Ghibli-esque setting
- + High source preservation of the characters' poses and clothing details
- − The facial features lean slightly towards generic shoujo manga rather than strictly Ghibli style
Wan 2.5 (Preview)
- + Strong Ghibli-specific facial character design and line work
- + Excellent soft lighting and atmospheric effects with floating leaves
- + Very accurate preservation of the original city background
- − The texture is a bit too smooth and digital, lacking the 'hand-painted' feel requested
- − The woman on the right has a slightly distorted arm connection
Verdict: Nano Banana Pro does a superior job of following the stylistic instructions, particularly with the hand-painted watercolor texture and the dreamy environment transformation. While Wan 2.5 captures the Ghibli character aesthetic very well, it feels more like a modern digital anime frame rather than the nostalgic, textured illustration requested.
Golden Hour Stroll
Image Editing“Add dynamic motion to this photo: make hair blow in the wind, add leaves flying, energetic and lively feel.”
AI Judge Analysis
Nano Banana Pro
- + Successfully added highly dynamic hair and tail movement.
- + The addition of autumn leaves creates a strong sense of wind direction and energy.
- − Excessive hair obscures the subject's face.
- − The dog's face underwent slight structural changes compared to the source.
Wan 2.5 (Preview)
- + Excellent source preservation of both the woman and dog's facial features.
- + Subtle, natural hair movement that doesn't obscure the subject.
- − Total energy is lower than requested; the leaves are sparse and look like flat overlays.
- − The dog's tail has a motion blur effect that looks a bit digital and unnatural.
Verdict: Nano Banana Pro followed the 'energetic and lively' instruction more aggressively by adding significant hair movement and a swirling leaf effect, though it sacrificed some facial clarity. Wan 2.5 (Preview) maintained much higher fidelity to the source image's identity, but its additions feel less integrated and the overall sense of motion is much weaker. Nano Banana Pro is preferred for better capturing the intended mood of the prompt.
Vintage Cafe Logo
Text-to-Image“Vintage minimalist restaurant logo for "Caffè Florian", retro cloche dome with steam and "Est. 1720" banner, classic typography, warm brown and cream tones, subtle texture on light background, vector emblem style.”
AI Judge Analysis
Nano Banana Pro
- + Perfect execution of the 'Caffè Florian' text with correct accent mark.
- + Strong vector emblem style with sophisticated cross-hatch shading and line work.
- + Clear, legible typography that feels high-end and professional.
- − The steam lines are somewhat detached from the cloche lid.
Wan 2.5 (Preview)
- + Good use of negative space for the steam effect inside the cloche.
- + Correct inclusion of all requested text elements within a banner structure.
- + The paper texture background is well-integrated.
- − Text rendering is inconsistent, with 'Est. 1720' being slightly warped.
- − The cloche design feels bulky and less elegant compared to Model A.
- − The composition feels a bit cramped within the central white square.
Verdict: Nano Banana Pro is the clear winner for its superior typography and sophisticated vector illustration style. While Wan 2.5 (Preview) produced a creative interpretation of steam inside the cloche, Nano Banana Pro delivered a more professional-looking logo with perfect spelling and better adherence to the 'vintage minimalist' aesthetic.
Apollo 11: Journey to Tranquility
Text-to-Image“Create a clean, modern vector infographic poster about the Apollo 11 mission. NASA-inspired palette (navy, white, muted red, light gray). Flat-vector style, crisp lines, consistent iconography, subtle gradients only. Steps (stop at landing): 1. Launch (Saturn Vicon) 2. Earth Orbit (Earth + orbit ring icon) 3. Translunar (trajectory arc icon) 4. Lunar Orbit (Moon + orbit ring icon) 5. Descent (lunar module descending icon) 6. Landing (lunar module on the surface icon) Small supporting elements (minimal text): • Crew strip: three silhouette icons with only last names: Armstrong, Aldrin, Collins. • Landing site marker: Moon pin labeled "Tranquility" only. Layout constraints: generous margins, large readable labels, clean background with subtle stars. Vector-only, print-poster look, high resolution.”
AI Judge Analysis
Nano Banana Pro
- + Excellent text legibility and correct spelling across all labels.
- + Matches the requested flat-vector style with professional, clean iconography.
- + Logical flow that clearly illustrates the 6 requested steps in sequence.
- − The rocket in the descent step is a bit stylized and less technical than the landing module.
Wan 2.5 (Preview)
- + Strong NASA-inspired color palette and crisp vector lines.
- + Clean illustration of the lunar landing module on the surface.
- − Confusing visual flow with overlapping lines and inconsistent icon placement.
- − Poor facial rendering in the crew section, especially for Michael Collins.
- − Misses the specific sequential numbering requested for the steps.
Verdict: Nano Banana Pro significantly outperforms Wan 2.5 (Preview) by delivering a coherent, logically organized infographic with perfect text rendering and adherence to the 6-step sequence. While Wan 2.5 has good individual illustrations, its layout is cluttered and the crew portraits are somewhat distorted.
Explore each model
Alibaba's text-to-image and image-to-image generation model from the Wan AI suite, offering high-quality visual generation capabilities