Nano Banana 2 Lite costs about 3 cents per image — 1,000 images for $30 — which works out to roughly one-quarter the price of Nano Banana Pro at volume (4× cheaper at 1,000 images) and half the cost of Nano Banana 2. It also generates about 2.5× faster than Nano Banana 2, so the saving compounds in time as well as dollars.
Start from the documented per-image numbers: Nano Banana 2 Lite generates at 3 cents per image, Nano Banana 2 costs twice that, and Nano Banana Pro runs roughly four times Lite's price at a 1,000-image volume. Scaling those documented rates linearly gives you the budget picture at any volume:
Volume | Nano Banana 2 Lite (~$0.03/image) | Nano Banana Pro (~4× Lite) | You keep |
|---|---|---|---|
1,000 images | $30 | ~$120 | ~$90 |
10,000 images | $300 | ~$1,200 | ~$900 |
100,000 images | $3,000 | ~$12,000 | ~$9,000 |
What the discount trades away. Lite outputs at 1K resolution — a deliberate design trade-off, not a quality failure. On raw text-to-image quality, the gap is smaller than the price suggests: per Google's own benchmarks, Nano Banana Lite outperforms Nano Banana Pro on text-to-image quality. The Pro premium buys resolution and the hardest, most technically demanding shots, not baseline image quality.
How to spend the difference. The two-tier workflow that makes the 4× multiplier pay: run all exploration, mockups, drafts, and concept testing on Lite, then escalate only selected assets to Nano Banana Pro for final production quality — as one review put it, "only the ones that are chosen get escalated to Nano Banana Pro." In practice this means a D2C brand can test 50 ad concept variants for less than the cost of a coffee instead of budget-limiting to 5, find the winner, and cut customer acquisition cost; agencies effectively eliminate internal draft-review generation costs; and filmmakers who treated 20 mood boards as a budget conversation now treat that exploration as practically free.
The real cost at scale isn't the model. Once generation is 3 cents and ~4 seconds, the bottleneck shifts from price to context — remembering what you told the model three images ago across a hundred-generation session. Tools like the invideo agent solve this with persistent memory: load your creative direction, character sheets, and brand context once, and every subsequent generation pulls from it automatically while the invideo agent routes exploratory work to Lite and chosen finals to Pro without you rebriefing per image. Do factor retries into your effective cost: if a Lite output needs regenerating for a final deliverable, that's a shot that belonged on Pro in the first place — the escalation split, not Lite-for-everything, is what preserves the savings.
Watch some of these to see what works for you:
if image generations is that cheap and that fast, you don't budget your generations anymore
— invideo's creative team