AI Image Generation API Cost Comparison
Enter kept images per month, takes per kept image and megapixels once. The calculator prices Recraft V4.1 Flash / V4.1 (raster, vector, pro raster) / V4, fal.ai FLUX 3, FLUX 2 / Flash, FLUX.1 dev / schnell, Z Image Turbo, Nano Banana / 2 / Pro and Seedream V4.5 / 5.0 Pro, Gemini API image models at 1K and OpenAI gpt-image (estimated at 1,500 output tokens per image), then sorts them by cost per kept image with billing unit and vector-output flags. Estimates only; official pricing pages prevail.
Math runs in your browser. Nothing is uploaded.
Cheapest per kept image
—
| # | Model | Billing | Vector | Per generation | Per kept image | Monthly |
|---|
Formula: per kept image = per generation × takes; monthly = kept × per kept. Per-megapixel fal rows scale with the megapixel field; per-image rows (Recraft, Nano Banana, Seedream, Gemini) do not. Gemini API rows use 1K standard (non-batch) rates (3 Pro Image 1K/2K share one price). OpenAI rows are an estimate at 1,500 image-output tokens per image, no image-input tokens, standard rates (gpt-image-2 / 2.5 $30 / 1M ≈ $0.045; gpt-image-1-mini $8 / 1M ≈ $0.012). Only Recraft returns native vector (SVG) output.
Estimates only. Official list prices as of 2026-11-01 (Asia/Shanghai); the provider's pricing page prevails. Sources: Recraft API pricing · fal.ai pricing · Gemini API pricing · OpenAI API pricing
Recraft API Cost Calculator · fal.ai Cost Calculator · Gemini Image Cost Calculator · OpenAI Image API Cost Calculator · AI Video Generation API Cost Comparison
More in this group
- Recraft API Cost Calculator Recraft raster/vector $/image → monthly + API units.
- fal.ai Cost Calculator fal image $/MP or $/image → per kept image & month.
- Gemini Image Cost Calculator Nano Banana images by resolution and Batch.
- OpenAI Image API Cost Calculator GPT Image tokens with optional Batch half-price.
- AI Video Generation API Cost Comparison 20+ video models ranked by $ per finished clip.
How it works
- Which AI image generation API is cheapest?
- At 1 megapixel per image, fal's FLUX.1 schnell ($0.003/MP) is cheapest, followed by Z Image Turbo and FLUX 2 Flash ($0.005/MP) and Recraft V4.1 Flash ($0.007/image). Nano Banana Pro ($0.15 on fal, $0.134 on Gemini 1K/2K) and Recraft V4.1 pro raster ($0.21) sit at the top.
- Why compare cost per kept image?
- Most usable images take more than one generation. Cost per kept image = per-generation cost × takes, so a cheap model you re-roll four times can cost more than a pricier model that lands first try.
- How are OpenAI gpt-image costs estimated?
- OpenAI bills image output by tokens. For a like-for-like row we assume 1,500 image-output tokens per image, no image-input tokens and standard (non-batch) rates: gpt-image-2 / 2.5 at $30 per 1M ≈ $0.045 per image, gpt-image-1-mini at $8 per 1M ≈ $0.012. Use the OpenAI image calculator for exact sizes and quality.
- Which APIs return vector (SVG) images?
- Among these, only Recraft returns native vector output — $0.08 per V4.1 vector image or $0.30 pro vector. Other models return raster images that would need a separate vectorize step (Recraft vectorize is $0.01).
- Do per-megapixel prices change with resolution?
- Yes. fal FLUX and Z Image rows multiply by megapixels, so 2 MP doubles their cost. Per-image rows — Recraft, Nano Banana, Seedream, Gemini at 1K — stay flat in this table.