Image Generation Cost Calculator
Price AI image generation per image and per month — gpt-image and Gemini rates, verified.
per image on gpt-image-2 (medium quality) — $6.34/day, $192.87/month at 200 images/day.
Monthly cost across models — same volume & tier
Rates verified 2026-06-11. Text prompt tokens (a few hundred per request) excluded — they add well under 1% at these rates.
How it works
This calculator prices AI image generation the way the invoices actually arrive: per image, scaled to daily and monthly volume. Pick a model and quality tier, set images per day, and the per-image rate, daily spend and monthly projection update instantly. Everything runs in your browser — no signup, no upload, no network call.
The pricing mechanics differ by vendor and that difference matters. OpenAI bills image models per million image output tokens — $30/M on gpt-image-2, $32/M on gpt-image-1.5 and $8/M on gpt-image-1-mini — and the quality tier sets how many tokens a 1024×1024 image consumes: roughly 272 at low, 1,056 at medium and 4,160 at high. That makes a high-quality render about 15× the cost of a draft on the same model. Google takes the opposite approach: Gemini 3 Pro Image is effectively flat per image, about $0.134 for 1K/2K output and $0.24 for 4K, billed as $120/M output image tokens.
The batch toggle models OpenAI's Batch API, which halves the rate for image models in exchange for up to 24-hour turnaround. For nightly product-catalog renders or thumbnail backfills, that is free money — the comparison bars show how the discount reorders the model ranking at your volume. Gemini has no batch discount, and the tool flags that rather than silently applying one.
Text prompt tokens are deliberately excluded: at $5/M (or $2.50/M on the mini model) a few-hundred-token prompt adds a fraction of a cent per image, well under 1% of the total. The numbers shown are the dominant cost, not a padded estimate. Monthly figures use a 30.44-day month (365.25 ÷ 12).
All rates were verified on 2026-06-11 against the OpenAI pricing page and Google's Gemini API pricing docs, and that date is embedded in the tool's source. Image pricing moves faster than text-token pricing, so the date is part of the answer. For measured spend from real traffic — including retries and failed generations this calculator cannot see — that is what FORG tracks automatically. The share link preserves your exact inputs for handing a scenario to a teammate.
Frequently asked questions
How is AI image generation priced?
OpenAI bills image models per million image output tokens — $30/M for gpt-image-2, $32/M for gpt-image-1.5, $8/M for gpt-image-1-mini — so per-image cost depends on quality tier, which sets the token count. Google's Gemini 3 Pro Image bills $120/M output image tokens, which works out to about $0.134 per 1K/2K image and $0.24 per 4K image.
What does quality tier change about the cost?
On OpenAI image models, a 1024×1024 image consumes roughly 272 output tokens at low quality, 1,056 at medium and 4,160 at high — so a high-quality image costs about 15× a low-quality draft on the same model. Gemini 3 Pro Image is flat-rate per image regardless of quality setting, with a separate higher rate only for 4K output.
Does the Batch API discount apply to image generation?
Yes, for OpenAI image models: batch requests are billed at 50% of the synchronous rate in exchange for a turnaround of up to 24 hours. That suits nightly catalog renders, thumbnail backfills and any pipeline where latency is irrelevant. Gemini 3 Pro Image does not currently offer a batch discount, so the calculator applies full rate there.
Are input or prompt tokens included in these numbers?
Text prompt tokens are billed separately ($5/M on gpt-image-2 and gpt-image-1.5, $2.50/M on the mini model) but a typical prompt is a few hundred tokens, which adds a fraction of a cent — well under 1% of the per-image cost. The calculator notes this exclusion explicitly rather than burying it, so the figure you see is the dominant cost driver.
How accurate are these rates?
Every rate was verified on 2026-06-11 against the OpenAI pricing page and Google's Gemini API pricing documentation, and the verification date is embedded in the tool itself. Vendors change image pricing more often than text pricing, so treat the date as part of the answer — if it is stale, re-check the vendor page before budgeting.
FORG tracks this automatically across every agent session — live cost attribution, budgets, and alerts.
Start tracking with FORG