MeiGen AI Gallery 移动端截图

Goal: Create a polished mobile webpage screenshot for an AI image prompt gallery called {argument name="gallery name" default="MeiGen Gallery"}, shown as a vertically scrollable feed of visual prompt examples.

Canvas: Tall smartphone screenshot, 720×1200 px, white and very pale blue background, rounded browser-like page edges, with a thin dark vertical scroll indicator on the far right near the top.

Header: At the top, place a minimalist black mountain-like triangular logo followed by the bold text “MeiGen” and a gray handwritten script word “Gallery”. Keep the header clean with generous white space.

Layout: Below the header, create a two-column masonry gallery inside a softly rounded light container. Use large rounded-rectangle image cards with consistent gutters and soft shadows. Show exactly 6 visible gallery cards:
1. Top-left large card: warm lifestyle portrait of {argument name="cafe subject" default="an elegant woman in a sleeveless cream blouse with a bow collar and lavender trousers"} seated at a round wooden cafe table beside tall industrial windows and a brick wall, holding a pencil near her lips, with an open book and iced coffee on the table; cozy cinematic sunlight.
2. Top-right large card: minimalist Chinese Dragon Boat Festival poster by water, vertical composition with a seated woman in white near a reflective pool, bamboo shadows on a pale wall, calm lake horizon, large Chinese characters “端午”, small English text “Slow into summer”, and subtle vertical Chinese typography.
3. Middle-left large card: energetic streetwear photo of {argument name="streetwear subject" default="a young man in a denim jacket, white shirt, beige shorts, white socks, and sneakers"} sitting on concrete steps in front of a bright blue graffiti mural with a cartoon gorilla, doodle arrows, crowns, X marks, and neon yellow accents.
4. Middle-right large card: snowy winter portrait of {argument name="winter subject" default="a woman wrapped in a thick blue patterned shawl"} standing outdoors at night or dusk, snowflakes visible, warm blurred lights in the background, moody cinematic depth of field.
5. Bottom-left partially visible wide card: clean product design/spec sheet collage with four labeled sections for futuristic gadgets: “AEROFOLD ONE” electric scooter, “HELIX CHEF” compact kitchen device, “SYNAPSE BAND” headphones or wearable band, and “LUMEN CAST” translucent tech device; include small renders, annotations, icons, and neat grid typography.
6. Bottom-right partially visible wide card: anime-style character banner featuring {argument name="anime character name" default="HINATA HYUGA"} in a white and purple action pose with long dark hair, purple UI accents, bold large text “HINATA HYUGA”, circular badges, and dynamic motion lines.

Visual style: The overall screenshot should look like a modern AI prompt inspiration website: crisp UI, soft rounded corners, realistic mixed media thumbnails, high-resolution images, balanced spacing, clean typography, and no clutter.

Constraints: Preserve the two-column gallery arrangement, show exactly 6 cards as listed, make the bottom two cards cropped by the lower edge as if the page continues, avoid extra cards, avoid watermarks, and keep all text legible where specified.