
Original document
GPT Image 2 · en
Scene:
Create a square premium Mid-Autumn campaign image at early evening with warm lantern-adjacent light, deep midnight blue, muted gold, lacquer red-brown, and soft ivory. The environment should feel refined and contemporary, with moon-window geometry or circular framing cues integrated into the space. Build a polished brand-world where gifting, tea culture, and family warmth are present through objects and atmosphere, not through overt festival spectacle.
Subject:
Center a premium mooncake and tea gift presentation with lacquer boxes, carefully plated pastries, brushed metal or ceramic tea elements, osmanthus-toned accents, and a round moon motif integrated into the composition. The objects should feel gift-ready, culturally grounded, and commercially elevated. Use layered foreground and background depth so the frame can function as a hero image for packaging, retail, or seasonal storytelling.
Important details:
Lock the Mid-Autumn grammar here: circular moon references, warm evening glow, refined lacquer finish, precise packaging geometry, elegant tea textures, and a balance of festivity with calm premium restraint. Avoid generic tourist aesthetics. Keep all surfaces believable and culturally respectful. The anchor should feel reusable across gifting campaigns, event rollouts, and retail surfaces while remaining unmistakably centered on Mid-Autumn generosity and reunion.
Use case:
Mid-Autumn series study 1/9.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no readable calligraphy, no zodiac animal mascot, no brand logos, no crowded banquet, no fantasy palace, no cartoon lantern overload, no fireworks, no kitsch souvenir styling.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.





