Prompt: Generate a Logo Concept
Generates one flat, vector-style logo concept for a brand, with the mark, colours and negative space specified instead of left to chance.
The prompt
Generate ONE logo concept image with the following exact specification.
Brand: {{BRAND NAME — e.g. Northbound Coffee Roasters}}, in the {{INDUSTRY — e.g. speciality coffee}} industry.
Concept: {{MARK CONCEPT — e.g. a minimalist mountain peak merged with a coffee bean silhouette}}.
Style: flat, vector-style logo mark, {{STYLE DETAIL — e.g. single-weight line work, geometric and modern}}, suitable for use at both business-card and shopfront-sign size.
Colours: {{COLOUR PALETTE — e.g. deep forest green and warm cream, two colours only}}.
Constraints:
- Exactly one mark, with the brand name set beneath it in a simple sans-serif typeface, not integrated into the mark itself
- Flat vector style only: no gradients, no drop shadows, no photographic texture, no 3D rendering
- Background: plain white, so the mark and its negative space are both clearly visible
- No tagline, no additional decorative elements beyond the mark and the brand name
Example: a two-colour geometric mark reads clearly even shrunk to the size of a favicon; a detailed illustrative one does not.
Return only the generated logo image, with no accompanying text or caption.
Before generating, check the concept, style and colour count above don't conflict (e.g. a "two colours only" instruction alongside a request for a gradient).
Example
Input
BRAND NAME — e.g. Northbound Coffee Roasters: Northbound Coffee Roasters INDUSTRY — e.g. speciality coffee: speciality coffee MARK CONCEPT — e.g. a minimalist mountain peak merged with a coffee bean silhouette: a minimalist mountain peak merged with a coffee bean silhouette STYLE DETAIL — e.g. single-weight line work, geometric and modern: single-weight line work, geometric and modern COLOUR PALETTE — e.g. deep forest green and warm cream, two colours only: deep forest green and warm cream, two colours only
Output (a written description, not a generated image)
What a good result should look like: A flat vector logo mark centred on a plain white background. The mark merges a minimalist mountain peak with a coffee bean silhouette, drawn in single-weight forest-green line work with a warm cream fill inside the peak's negative space, so the bean shape reads clearly within the mountain's outline. The lines are geometric and evenly weighted throughout, with no gradients, shadows or photographic texture anywhere in the image. Beneath the mark, "Northbound Coffee Roasters" appears in a simple, modern sans-serif typeface, set in the same forest green and clearly separated from the mark above it. No tagline or decorative elements appear. The mark is compact and legible enough that it would still read clearly if scaled down to a small size, consistent with a two-colour geometric logo built for both signage and small digital use.
Why it works
- Separating the mark concept from the colour and style constraints controls whether the logo looks distinctive or generic, more precisely than a single vague brief.
- Explicitly banning gradients, shadows and 3D rendering keeps the output usable as a real vector-style mark rather than an unscalable illustrated image.
- The pre-generation consistency check catches a self-contradictory brief, such as a two-colour rule alongside a gradient request, before generation.
Why the basic version fails
“Design a logo for my coffee brand” leaves the mark concept, the colour count and the style entirely open, so you typically get an overly detailed, photorealistic or gradient-heavy image that looks impressive as a picture but doesn’t work as an actual logo — it won’t scale down, reproduce in one colour, or read clearly as a small icon.
What this prompt fixes
Fixing the mark concept, the exact colour count and a flat vector style is what actually makes an output usable as a logo rather than just a nice image. Banning gradients, shadows and photographic texture forces a result that would still reproduce correctly on a sign, a favicon or an embroidered cap, which a heavily rendered image will not.
Tips per model
- Gemini responds well to an explicit “flat vector style” instruction; without it, results often drift toward a more illustrated or 3D look.
- ChatGPT’s image generation follows an explicit colour count (“two colours only”) more reliably when it’s stated as a hard constraint rather than a stylistic suggestion.
- Treat this as a concept sketch, not a final production file — a usable logo still needs manual vectorising and refinement by a designer before real-world use.