Gemini image models: cost per 100 images at 1K, standard vs batch

At 1K, Google's Gemini API lists $0.0336 per image for Nano Banana 2 Lite, $0.067 for Nano Banana 2 and $0.134 for Nano Banana Pro. For 100 images that is $3.36, $6.70 and $13.40 on the standard tier, and half of that on the batch tier.

Batch is asynchronous, so it fits overnight catalog work, not an interactive tool.

Per 100 images at 1K (read 2026-10-04)

Batch is listed as 50% of standard on Google's pricing page, so the batch column is half the standard column.

Other sizes (read 2026-10-04)

Nano Banana 2 is listed at $0.045 for 0.5K, $0.101 for 2K and $0.151 for 4K. Nano Banana Pro is $0.134 at 1K or 2K and $0.24 at 4K.

What Sume charges

These are Google's list prices. Sume's price for a request is the cost_usd pricing line on the model, and the Image API also lists google/nano-banana-2 and google/nano-banana-pro with resolution tiers 512, 1K, 2K and 4K.

prices = {"lite": 0.0336, "nb2": 0.067, "pro": 0.134}
for name, price in prices.items():
    print(name, round(price * 100, 2), round(price * 50, 2))

When batch is the right call

Use batch when nobody is waiting: overnight catalog refreshes, a seasonal product set, a backlog. Use the standard tier when a person is staring at the screen.