AI Image Cost Per Batch Calculator
Estimate the cost of a batch of AI-generated images from the price per image, the number of variations and how many prompts you run.
How it works
Images = prompts × images per prompt · Total cost = images × price per image · Cost per usable = total ÷ (images × keep rate)
The headline cost is just the number of images generated times the per-image price, but the number that matters for planning is the cost per image you actually keep. Since generative models produce several variations per prompt and you discard most of them, the true unit cost of a usable asset is far higher than the sticker price per generation. Lowering how many variations you request, or raising your keep rate with better prompts, both cut the effective cost.
Worked example
At $0.04 per image, 50 prompts with 4 variations each generate 200 images for $8.00. If you keep 25% of them, that is 50 usable images, so each usable image effectively costs 8.00 ÷ 50 = $0.16.
Frequently asked questions
Why is my cost per usable image so much higher than the price per generation?
Because you pay for every image generated but only keep a fraction. If you keep one in four, your usable images cost four times the per-generation price. Improving prompts to raise the keep rate is often cheaper than generating more variations.
How do I lower the batch cost?
Generate fewer variations per prompt, use a lower-resolution or cheaper model for drafts, and refine prompts before scaling up. Reserving the expensive high-resolution runs for prompts you have already validated avoids paying premium rates for throwaways.
Does resolution or steps affect the price?
Usually yes — higher resolutions, more diffusion steps or upscaling passes cost more per image on most platforms. Enter the price for the exact settings you plan to run, and price draft and final passes separately if they differ.
Related calculators
This calculator is for educational purposes. Double-check important results before acting on them.