Free guide · works with ChatGPT, Claude or Gemini
Chatgpt prompts for image editing
Transform any raw photo into a clean, on-brand asset with correct sizing, filenames, and alt text in under 15 minutes once setup is complete. Includes a reliable workflow that creates auditable artefacts and reusable templates tailored to your chosen editor.
Step 1, free.
Copy this, paste it into your AI, and swap the [BRACKETS] for your own details.
Setup: Define brand and channel profile
Provide: [BRAND OVERVIEW], [TONE & STYLE], [IMAGE USE-CASES], [CHANNELS], [PRIMARY EDITOR] (e.g., Photoshop, Affinity, Figma), [COLOUR VALUES] (HEX/RGB/CMYK), [TYPEFACES], [LOGO ASSETS LINK], [IMAGE DIMENSIONS PER CHANNEL], [ACCESSIBILITY NOTES], [EXAMPLE LINKS], [CONTACTS], [APPROVAL SLA]. Act as a senior brand producer and distil a practical, unambiguous Brand & Channel Profile for image production. Work through it like this: 1) Reconcile brand identity into a controlled vocabulary (5–9 adjectives) and imagery do/don’ts. 2) For each channel, define target sizes, safe areas, focal zones, aspect ratios, and compression targets. 3) Lock colour usage: primaries/secondaries, neutrals, and allowed photo treatments (tinting, contrast ranges, grain). 4) Lock typography-in-image: type scales, weights, legibility thresholds, minimum sizes, and logo clear-space. 5) Define file standards: working colour space, bit depth, base canvas, grid, and default export settings per channel. 6) Note accessibility: colour contrast targets, text-in-image limits, and alt-text style. Give me the output as: - Brand & Channel Profile (one page) with sections: Identity, Imagery Do/Don’ts, Channels Table (Channel | Use-case | Dimensions | Aspect | Safe Areas | File Type | Compression), Colour & Treatment, Typography & Logo, Accessibility, File Standards, Contacts & Approvals, Editor Defaults. Quality bar: Everything is specific and measurable; no stylistic fluff; conflicts are resolved with a single stated rule per case. If information is missing or unclear, mark MISSING and ask concise follow-up questions grouped at the end.
How to get a good answer.
- 01Fill every [BRACKET] with real detail. Vague in, vague out.
- 02Read the answer, then reply with what's wrong or missing. The AI fixes it in seconds.
- 03Check facts, prices and anything legal or safety-related yourself before you use it.
The rest of the workflow.
Step 1 gets you started. The full Polished Images in 15 Minutes takes you all the way, 14 steps in order:
- 01Setup: Define brand and channel profile (free above)
- 02Setup: Create image intake and file hygiene
- 03Setup: Draft one-page Edit Brief
- 04Production: Audit the image
- 05Production: Set edit goals and constraints
- 06Production: Build the prompted edit script
- 07Production: Execute non-destructive edits
- 08Production: Version and compare candidates
- 09Quality Control: Run visual and accessibility checklist
- 10Quality Control: Stakeholder review and change log
- 11Delivery: Finalize export specs and filenames
- 12Delivery: Generate alt text and captions
- 13Delivery: Package and deliver assets
- 14Maintenance & Reuse: Archive, template, and reuse loop
Pay once, keep it. Delivered instantly with a copy button on every prompt.
Questions
Does this work with the free version of ChatGPT?
Yes. The prompt works in the free versions of ChatGPT, Claude and Gemini, on a phone or a computer.
What do I put in the [BRACKETS]?
Your real details: notes, numbers, names, whatever the bracket asks for. The more specific you are, the better the answer.
Is the AI's answer always right?
No. Treat it as a strong first draft: check facts, figures and anything legal or safety-related before you use it.
What's in the full Polished Images in 15 Minutes?
14 steps run in order, a system prompt that sets the AI up for the job, worked examples, fix-it prompts and a quality checklist. It's NZ$29.50 once, no subscription.
Know someone who'd use this?
Send them the free prompt: