Master prompt
GPT-Image-1 — Singapore COMPASS scorecard infographic
Service-specific visual metaphor for Singapore Employment Pass: an illustrated 4-criterion COMPASS scorecard (C1 salary, C2 qualifications, C3 diversity, C4 local employment) with point breakdown.
ImageGPT-Image-1SGMarketingInfographicCOMPASS
**TARGET MODEL: ChatGPT Images 2.0 (gpt-image-2, April 2026 release).** Use 2.0-specific capabilities:
- Embed exact text strings — 99% accuracy English, 90%+ Hindi/Punjabi/Bengali/Arabic/CJK. Spell every word verbatim in the prompt; the model will render it as typed.
- Let 2.0 plan composition first (the model uses an internal reasoning step). Specify spatial layout in plain language ("primary subject left-third, accent right-third, headline top-center").
- Request 2K (2048x2048 or 2048x1152) when the output is for web hero / OG / LinkedIn — use 1024px square only for thumbnails.
- For carousel / slide-deck use cases, ask for a coherent SET of up to 8 images in one call.
- If multilingual diaspora copy is needed (Punjabi for Brampton, Hindi for North-India market, Tamil for SG diaspora), specify the EXACT script + transliteration to lock the rendering.
---
Generate a Singapore COMPASS scorecard infographic at 1:1 (1024x1024 px target).
Composition:
- Background off-white #F6F4EF, 2% paper-grain.
- Top 12%: eyebrow "SINGAPORE EMPLOYMENT PASS" small caps Geist sans #1E3A5F 70%, headline "COMPASS - the 4 scoring foundations" Fraunces serif #1E3A5F.
- Center 75%: a 2x2 grid of equal-sized cards, separated by thin off-white #E5E7EB rules.
Top-left card "C1 - Salary": label small caps #5E54FF top, "Max 20 pts" Geist Mono right, body "Salary relative to top third of local PMET sector benchmark" Geist sans 11pt #1E3A5F. A small abstract bar-glyph indicator showing 20 points filled.
Top-right card "C2 - Qualifications": "Max 20 pts", "Top 100 university gets full points, recognised institutions get partial". Bar-glyph 20.
Bottom-left card "C3 - Diversity": "Max 20 pts", "Nationality complementarity bonus - rewards firms whose foreign workforce is mixed". Bar-glyph 20.
Bottom-right card "C4 - Local employment": "Max 20 pts", "Local PMET headcount above industry average". Bar-glyph 20.
Center of grid: a small "40 PASS / 80 EXCEED" pill badge in #F28C28.
- Bottom 13%: one-line takeaway Fraunces italic #1E3A5F: "40 base points to pass. Bonus C5/C6 criteria can lift borderline cases."
- Bottom-left: "[FIRM_NAME]" tiny serif italic 50%.
- Bottom-right: "DRAFT" stamp #F28C28 rotated -12 degrees ~50px.
Color palette: off-white, navy, indigo, draft orange.
Typography: Geist sans for body, Geist Mono for points, Fraunces serif for headline + takeaway. Render each criterion code literally including the hyphen.
Style: Editorial 2x2 grid infographic. Clean, flat, no 3D.
Negatives:
- No MOM logo, ICA logo, Singapore coat of arms.
- No Singapore flag.
- No stock-photo people.
- No fake/invented point values - 20 per criterion is correct.
GPT-Image-1 tips: state grid order top-left, top-right, bottom-left, bottom-right. State each criterion code literally.
DRAFT - for consultant/firm brand review. Verify generated image does NOT include trademarked logos, real-person likenesses, or any government-issued seal that could be mistaken for an official document. Brand colors must align with firm guidelines. Generated images are for marketing only - never use to imply official endorsement.Unlock the vault to see the full prompt
