Image Prompt Generator
Detailed prompts for any image AI tool
Generate rich, optimized prompts for Midjourney, DALL-E, Stable Diffusion, and other image generation tools from a simple concept description — including style references, lighting, composition, and negative prompt recommendations.
GPT-6 Luna
for 100K tokens a month · 40% input / 60% output
Why this model
For this workload — 100K tokens a month, 40% of them input — it comes to about $0.03 a month, the lowest of the models suggested here.
Alternative models
Implementation tips
- 1
Build model-specific prompt templates: Midjourney responds to aspect ratios and style parameters, DALL-E to natural language descriptions, and Stable Diffusion to weighted token lists — a universal approach underperforms on all three.
- 2
Generate 5 prompt variants for each concept at different aesthetic directions (photorealistic, painterly, minimal, cinematic, editorial) — creative teams rarely know which direction works until they see options.
- 3
Include a 'negative prompt' section in your output for models that support it: listing what to avoid (blurry, distorted hands, watermarks) dramatically improves output quality on the first generation.