Anime Model Comparison Dashboard

Goal: Create a dark, widescreen AI image-comparison interface showing one source prompt card and three generated-image result cards, like a model comparison board for anime artwork.

Canvas: 16:7 horizontal screenshot-style composition on a nearly black background with faint dotted grid texture and subtle diagonal guide lines. Use soft glow, rounded panels, thin gray borders, and a polished dark UI aesthetic.

Layout: Arrange exactly 4 rounded rectangular cards in one row from left to right. Card 1 is a narrow text prompt card; cards 2, 3, and 4 are larger vertical image cards. Leave small gaps between cards. Each card has a tiny header line at the top with a small icon and Japanese-style label. Add a small gold coin-like icon near the top-right of the first card area. Add small circular connector or action buttons between the cards, aligned vertically around the middle and lower edge.

Card count and contents: Exactly 4 visible cards total: 1 text card and 3 image result cards. The 3 image result cards each show the same subject rendered slightly differently: an anime-style young woman with very long silver-white hair and bangs covering her eyes, pale skin, a black cropped sleeveless top, black pants, and an oversized loose gray jacket lifted or held behind her shoulders. Her pose is confident and fashion-editorial, with arms raised behind her head or jacket spread open. The palette is desaturated gray, white, black, and muted beige.

Text card: The first card header reads approximately “テキスト #9”. Inside it, include a dense English prompt in small white text describing an anime close-up portrait: “Medium close-up to three-quarter body portrait, slightly high camera angle, camera positioned just above eye level, mild perspective compression, fashion editorial framing, character occupying most of the frame, clean warm-white studio background...” Continue with compact prompt-style text about contemporary Japanese anime illustration, thin line art, refined facial proportions, silver white hair, high crop black top, oversized jacket, muted/desaturated colors, soft shadows, vignette, premium poster presentation. Keep text small but legible-looking, not necessarily fully readable.

Image card headers: Label the three result cards with Japanese-style headers: “画像生成ツール #33”, “画像生成ツール #34”, and “画像生成ツール #35”. At the bottom of each image card, add a faint Japanese status caption in white/gray, similar to “認証されたプロンプト、@を使って参照や追加のコンテキストを加えられます”.

Subject details: The character has {argument name="hair color" default="silver-white"} very long hair with heavy bangs obscuring the eyes, a slim athletic waist, pale skin, and a moody expression mostly hidden by hair. Outfit consists of exactly 4 main pieces: black cropped high-neck sleeveless top, oversized gray jacket, black low-rise pants, and dark belt. Render the images in a {argument name="anime style" default="dark fantasy contemporary anime"} look with clean yet painterly detail.

Visual style: Dark fantasy anime fashion, muted monochrome palette, soft studio lighting, editorial composition, slight vignette, high contrast between white hair and black clothing, crisp UI screenshot presentation. The first generated image should look sharp and classic anime, the second slightly more painterly and shadowed, the third cleaner and more minimal with a bright pale background.

Constraints: Show exactly 4 cards and exactly 3 anime result images. Do not add extra characters, logos, watermarks, browser chrome, or unrelated UI. Keep the overall image as a comparison dashboard rather than a single illustration.