
Original document
GPT Image 2 · en
Scene:
A square photorealistic overhead photograph of a fictional herbal study note on warm cream prescription-style paper, placed beside a ceramic tea cup, blank kraft envelope, and a brush pen. The paper has slight wrinkles, deckled fiber edges, and a red specimen chop mark. Lighting is soft and domestic, like a careful product shot for cultural stationery.
Subject:
Create a fictional traditional-herbal note, explicitly not a medical prescription. The exact readable text should include: SPECIMEN, Herbal Study Note, Not for treatment, 甘草, 陈皮, 茯苓, 菊花, 生姜, sample only. Mix tidy printed form fields with expressive handwritten Chinese herb names, using dark brown ink and a red square stamp reading SPECIMEN. Do not include patient name, doctor name, diagnosis, dosage, pharmacy, license number, or instructions for use.
Important details:
The visual goal is document-level precision: ruled boxes, date area left blank, blank signature line, ink pressure changes, brush-pen pooling, paper translucency, red stamp texture, and believable stationery shadows. The Chinese characters should be readable and beautiful, but the content must remain a non-actionable study sample. Keep the note materially convincing while making its specimen and non-treatment status unambiguous.
Use case:
Pseudo-Real Documents single study 3/8, testing handwritten document realism, Chinese character rendering, paper fibers, and safe non-medical framing.
Constraints:
No readable text except the specified herbal study strings, no watermark, no logos, preserve exact 1:1 aspect ratio, no patient data, no doctor signature, no dosage, no treatment directions, no real clinic name, no pharmacy seal, no scannable code.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.





