请整理这段画面创意的生成提示词,保留主体、构图、光线与风格细节。
复古现代主义电影感插画,主体为[主体]。大面积高纯度色块构成画面,深青绿色背景、红橙色建筑平面、奶油白地面,强烈冷暖互补色。极简几何构图,建筑切面感,硬光侧照,长而清晰的深色投影。主体安静、孤独、沉思,画面留白充足,氛围静默、神秘、诗意,像被阳光凝固的午后。平涂色块结合细密线条,局部细节精致,轻微颗粒、哑光印刷、复古纸张质感。无文字,无水印,无 logo,无 3D,无摄影感。
Send text messages and continue the conversation. Output is limited to 4096 tokens per reply.
Try this model
Model inputs
Estimated credits: 1+
Chat controls
Choose a reasoning level, turn web search on or off, and decide whether to show model thoughts. Available controls depend on the selected model.
Thoughts
Use the model default or choose whether to show its thought output, when supported.
Ask a question or describe a task to start.
Model examples
Review examples to help decide whether the model fits your task.
请整理这段画面创意的生成提示词,保留主体、构图、光线与风格细节。
复古现代主义电影感插画,主体为[主体]。大面积高纯度色块构成画面,深青绿色背景、红橙色建筑平面、奶油白地面,强烈冷暖互补色。极简几何构图,建筑切面感,硬光侧照,长而清晰的深色投影。主体安静、孤独、沉思,画面留白充足,氛围静默、神秘、诗意,像被阳光凝固的午后。平涂色块结合细密线条,局部细节精致,轻微颗粒、哑光印刷、复古纸张质感。无文字,无水印,无 logo,无 3D,无摄影感。
请整理这段画面创意的生成提示词,保留主体、构图、光线与风格细节。
I’m uploading a character/PFP reference image. Your task is to completely reimagine this character in the style of a high-end cyberpunk anime fashion illustration similar to a luxury Japanese streetwear poster. The uploaded image is BOTH the identity reference and the clothing/fashion inspiration reference. Preserve the core facial identity, recognizable hairstyle, facial proportions, silhouette, personality, anime identity, and overall fashion energy of the original character. The new outfit must feel directly inspired by the uploaded character’s original clothing design, including the jacket shape, layered styling, color palette, Japanese streetwear influence, rebellious anime aesthetic, silhouette language, graphic placement style, and overall fashion attitude. Do not copy the outfit 1:1, do not create unrelated futuristic clothing, and do not replace the fashion with generic techwear. Instead, create an elevated luxury-editorial reinterpretation of the original outfit that feels fashionable first and cyberpunk second. The outfit should feel like a premium futuristic redesign by an elite Japanese fashion concept artist, featuring designer-inspired construction, luxury fabric layering, futuristic tailoring, elegant straps and hardware, subtle cyberpunk detailing, and premium sneaker aesthetics while remaining unmistakably connected to the uploaded character’s original fit. The image should feel clean, minimal, cinematic, stylish, expensive, fashion-forward, and highly polished. Prioritize simplicity, negative space, elegant color blocking, strong silhouette readability, premium composition, and soft painterly rendering. The rendering style should be painterly anime illustration with semi-realistic anime rendering, soft cel-shaded hybrid shading, subtle brush textures, atmospheric gradients, smooth ambient shading, minimal linework, elegant edge control, and polished anime poster quality. Use soft gradient shadows, colored shadows instead of black, atmospheric bounce lighting, subtle reflected light, soft contact shadows, and diffused studio lighting. The character must be shown full body from head to toe, standing naturally in a relaxed fashion-editorial pose, facing toward the camera with subtle eye contact, centered in frame with relaxed posture and luxury campaign energy. Do not generate a side profile, action pose, combat pose, anime attack pose, cropped body, waist-up framing, exaggerated motion, or extreme perspective distortion. Use an eye-level camera angle with full-body fashion photography framing, minimalist editorial composition, balanced negative space, clean silhouette readability, and poster-like symmetry. The composition should feature a minimalist background with a dominant magenta, pink, or red flat-color backdrop, subtle gradients, soft shadows behind the character, simple graphic design elements, and a subtle circular graphic shape behind the subject. Avoid detailed cityscapes, cluttered cyberpunk environments, tactical armor aesthetics, and heavy sci-fi machinery. Use a controlled color palette of hot pinks, magentas, soft cream whites, coral gradients, dusty rose shadows, muted warm skin tones, and subtle mauve shading with vaporwave-inspired harmony. The final image should feel like a premium anime fashion campaign illustration created by a top-tier AAA concept artist for a luxury Japanese streetwear brand while remaining unmistakably connected to the uploaded character’s original outfit and identity. Aspect ratio 3:4.
Model strengths
Model strengths
Suggested uses
Tasks to consider
Claude Opus 4.5 FAQ
Send text messages and continue the conversation. Output is limited to 4096 tokens per reply.
Credits are reserved before submission. Final billing uses confirmed consumption; any unused reserve is returned after settlement. If submission or consumption cannot be confirmed, the reserve remains held for reconciliation. Check credit history before sending again.
Compare alternatives
Work through pasted text with Claude: ask for summaries, compare proposals, or revise a draft through follow-up chat.
Why consider it
Credit charges: Base rate per million tokens: 400 credits for input, 2,000 for output
Use Claude Sonnet 5 on Rivya to discuss pasted requirements, draft implementation briefs, and prepare technical handoffs.
Why consider it
The final cost is rounded up to whole credits.
Example output
Claude Opus 4.5
Chat
Example output
Claude Opus 4.5
请整理这段画面创意的生成提示词,保留主体、构图、光线与风格细节。
复古现代主义电影感插画,主体为[主体]。大面积高纯度色块构成画面,深青绿色背景、红橙色建筑平面、奶油白地面,强烈冷暖互补色。极简几何构图,建筑切面感,硬光侧照,长而清晰的深色投影。主体安静、孤独、沉思,画面留白充足,氛围静默、神秘、诗意,像被阳光凝固的午后。平涂色块结合细密线条,局部细节精致,轻微颗粒、哑光印刷、复古纸张质感。无文字,无水印,无 logo,无 3D,无摄影感。
Best for
Supports
Logs show a config change at 10:02, errors at 10:04, alert at 10:09, rollback at 10:13 and recovery at 10:16. Write a neutral incident note.
The service began returning elevated errors two minutes after a configuration change. The alert fired at 10:09; the operator reverted the change at 10:13, and the error rate returned to baseline at 10:16. The observed impact window was approximately 12 minutes. Timing suggests the change may be involved, but the cause remains to be verified. Follow-ups: compare the two configurations and review the five-minute detection delay.
请整理这段画面创意的生成提示词,保留主体、构图、光线与风格细节。
[BRAND NAME]. Act as a Senior Industrial Designer and Technical Illustrator, specializing in pencil engineering sketch reproductions of brand logos (style: hand-drawn technical blueprint/concept sketch on aged paper). PHASE 1: AGED PAPER SURFACE A warm, aged beige-cream paper surface (#E8D5B0 range) filling the entire frame — with visible aging spots, worn patches, scattered pencil smudges, and fine paper grain consistent with a real old technical sketchbook or archival drawing sheet. Subtle vignette darkening at all frame edges. No clean digital white — the paper must read as genuinely old and physical. PHASE 2: LOGO RECONSTRUCTED AS PENCIL TECHNICAL SKETCH (TRUE SHAPE) [BRAND NAME]'s real, official primary icon/logomark is reproduced with full shape and proportional accuracy exactly as it appears in the brand's real guidelines, then rendered entirely as a hand-drawn pencil engineering sketch — as if a designer sketched the logo concept by hand in a technical drawing session. The logo is positioned in the center of the frame, occupying roughly 40–50% of frame height. CRITICAL ACCURACY INSTRUCTION — NON-NEGOTIABLE: Do not invent, approximate, restructure, or substitute the logo with any other imagery. The logo's real geometry and proportions stay exactly as in the brand's real guidelines — only the pencil sketch rendering treatment is applied on top of the unchanged shape. PHASE 3: PENCIL SKETCH TECHNIQUE (HERO EFFECT) Two distinct pencil layers build the complete illustration: - CONSTRUCTION LAYER (light pressure): thin, precise straight lines forming a rectangular bounding box around the logo's outer dimensions, plus internal diagonal construction/guide lines radiating from key structural points of the logo — all drawn with light pencil pressure, slightly imprecise, with the natural slight waviness of a hand-drawn straight line. - LOGO LAYER (heavier pressure): the logo's actual forms drawn with firmer pencil pressure — clean but hand-drawn contour lines tracing the real logo shape, with dense diagonal crosshatch pencil shading filling the solid/dark areas of the logo to imply volume and depth. Shading direction is consistent (45-degree diagonal strokes), varying in density from light to dark across the form. PHASE 4: TECHNICAL ANNOTATIONS Around the logo, hand-drawn engineering dimension annotations in classic technical drawing style — rendered in thin pencil handwriting: - Horizontal and vertical double-arrow dimension lines with end caps, indicating key measurements (written as plausible dimensions: "L=Xmm", "R=Xmm") - Angle annotations for key diagonal elements (e.g. "10°", "OFFSET X°") - All annotation text in a clean, slightly imperfect hand-lettered technical style — not a printed font, but the careful handwriting of a technical draftsman - 4–6 annotation labels total, positioned logically around the bounding box and key logo elements PHASE 5: COMPOSITION Logo and annotations centered in the frame with generous aged-paper negative space surrounding them, especially above and below. No color, no additional graphic elements — strictly monochrome pencil on aged paper. TECH SPECS: Flat 2D pencil illustration, no 3D, no digital gradients — all tone from pencil pressure and crosshatch density. Monochrome: dark graphite pencil on warm aged paper. Resolution: 8K.
Explain why 1/2 equals 3/6 to a child using a simple example.
Imagine one chocolate bar split into two equal pieces. Taking one piece gives you half the bar. Now split each of those two pieces into three smaller equal pieces. The bar has six small pieces, and your half contains three. The amount of chocolate has not changed: one half is three sixths.
What should the owner confirm before acting?
Confirm that the six pieces in the drawing are equal. Counting unequal pieces would not demonstrate the fraction equivalence.
We are investigating why teams stop using a shared calendar. Write five neutral questions.
1. Tell me about the last time your team used the calendar. 2. What happened between creating an event and attending it? 3. Where do you now record the same information? 4. Which part required the most effort? 5. What changed when the team stopped using it? Ask for concrete examples before suggesting possible reasons.
Model details
Provider
Anthropic
Category
Chat
Capabilities
Chat
Credit charges
Base rate per million tokens: 285 credits for input, 1,430 for output
Inputs
Prompt only
Prompt requirements
Up to 20,000 characters
Developer access
Use Claude Opus 4.5 through Public API v1. Check the model reference for supported inputs, parameters, and credit charges before sending a request.
Credit charges: Base rate per million tokens: 170 credits for input, 855 for output
A text-only Claude model for online chat on Rivya. Paste a draft or a pair of proposals and ask for a rewrite, comparison, or list of questions.
Why consider it
Credit charges: Base rate per million tokens: 400 credits for input, 2,000 for output