The cheapest image generation API depends on the output resolution, so the same model does not always rank the same way. As of September 2026, GPT-Image-2 costs $0.0633 per image at 1K, 2K, and 4K, while Google's models on the same key have separate prices for each resolution tier. As a result, GPT-Image-2 is the most expensive generation model at 1K and the cheapest at 4K — below both Nano Banana 2 at $0.1057 and Nano Banana Pro at $0.1200. The rankings flip within the range that comparison articles often collapse into a single number. Choose the output resolution you actually need before choosing a model.
First, a disclosure: Kunavo is our product, a gateway that resells access to models, not an independent benchmark. The figures below are read when the page is rendered from the catalog that actually bills requests. Final verification of Kunavo rates: October 4, 2026.
Price per image, by model and resolution
One price per image — the unit is the output, not the prompt length. “Provider list price” is the price the provider publishes for the same image; for Google, see the Gemini API image generation documentation, and for OpenAI, see the API pricing page. A dash means the model does not have a separate tier for that resolution.
| Model | Provider | 1K | 2K | 4K | Provider list price (standard) |
|---|---|---|---|---|---|
| Nano Banana 2 Lite | $0.0255 | — | — | $0.0340 | |
| Nano Banana | $0.0273 | — | — | $0.0390 | |
| Nano Banana 2 | $0.0469 | $0.0707 | $0.1057 | $0.0670 | |
| GPT-Image-2 | OpenAI | $0.0633 | $0.0633 | $0.0633 | $0.1266 |
| Nano Banana Pro | $0.0670 | — | $0.1200 | $0.1340 |
Models without separate resolution tiers are billed at a single rate regardless of the requested size. Every rate in the table is below the provider's published list price for the same image.
Models billed by rendering speed rather than resolution
Ideogram v3 does not have resolution tiers. It is billed by rendering speed — the amount of computation spent on each image — which works similarly to resolution. The cheaper setting is for drafts; the more expensive setting is for a final image. Keep this in mind before comparing the figures in the table above with this model's typical rate: the two figures measure different things.
| Model | Rendering speed | Per image | Official provider list price |
|---|---|---|---|
| Ideogram v3 | Turbo | $0.02625 | $0.0300 |
| Ideogram v3 | Balanced | $0.0525 | $0.0600 |
| Ideogram v3 | Quality | $0.0750 | $0.0900 |
Every tier is below Ideogram's own published price for the same setting. Ideogram v3 Edit is billed at the same 3 speed levels and is covered in the editing section below — it is the only model on this key that accepts masks.
Where the rankings flip
At 1K, comparison articles have it right: the smaller Google model is the cheapest, and GPT-Image-2 is among the more expensive. At 4K, the order flips. OpenAI does not charge extra for larger outputs, while Google does. As of September 2026, the figures on Kunavo are:
- At 1K — Nano Banana 2 Lite $0.0255, GPT-Image-2 $0.0633. GPT-Image-2 is about 2.5 times the price of the cheapest option.
- At 4K — GPT-Image-2 $0.0633, Nano Banana 2 $0.1057, Nano Banana Pro $0.1200. GPT-Image-2 is now the cheapest of the three.
For a product that generates thumbnails, Google models have a major cost advantage. For print-resolution assets, the single-rate model has the advantage. Articles that name a single winner without explaining which of these use cases they assume are leaving out important context. To calculate costs using your actual monthly image volume, use the AI Image Cost Calculator — changing the resolution tier shows how the model rankings change.
When editing existing images
There are 3 editing models that accept images as input. They use the same key and balance.
| Model | Per image | Multiple image inputs | Mask inpainting |
|---|---|---|---|
| Nano Banana Edit | $0.0273 | Supported | Not supported |
| GPT-Image-2 Edit | $0.0633 | Supported | Not supported |
| Ideogram v3 Edit | From $0.02625 (by speed) | Not supported | Supported |
Models that accept multiple source images in a single request: Nano Banana Edit and GPT-Image-2 Edit. This capability is needed for compositing, and edits are directed through prompts rather than painted regions. Ideogram v3 Edit is the only model that accepts mask-based inpainting (since September 14, 2026). Choose it for workflows such as retouching, where you paint the area to change; choose the other 2 models when compositing multiple images. Each edit is also billed at the price of one image, so a pipeline that generates an image and then edits it twice costs the equivalent of three images per final result.
What Kunavo does not offer
The 5 model families that often appear on shortlists for this search are not available on Kunavo. As of September 2026, these are Flux, Seedream, Recraft, Stable Diffusion, and Midjourney. If your shortlist centers on one of these, another provider is the right choice; this page is not trying to persuade you to switch from them. Ideogram appeared on this list until September 14, 2026, but it is now available with this key — see the rendering speed section above.
How to make a call
Generation and editing each use a single endpoint, both in an OpenAI-compatible format. The only thing that changes between models is the model ID. Unlike video, image generation returns results immediately without task polling. Request format and sizes are covered in the image endpoint documentation, with model-specific details in the Nano Banana guide and GPT-Image-2 guide. For video on the same key, see video generation API comparison.
Frequently asked questions
Which image generation API is the cheapest?
The answer depends on the output resolution. As of September 2026, on Kunavo, Nano Banana 2 Lite is the cheapest at 1K, at $0.0255 per image, while GPT-Image-2 is among the more expensive at $0.0633. At 4K, however, GPT-Image-2 costs $0.0633, the same as at 1K, while Nano Banana 2 rises to $0.1057 and Nano Banana Pro to $0.1200, making GPT-Image-2 the cheapest of the three. OpenAI does not charge more for higher resolutions, while Google does. Choose the resolution before choosing the model.
How much does image generation API pricing cost?
As of September 2026, prices on Kunavo range from $0.0255 to $0.1200 per image. Billing is per output image, not by token count, so a token calculator will not give you the cost of this task. Charges are deducted from your prepaid balance as you use the service. Failed generations are not billed, and there is no monthly minimum.
Does Kunavo offer Flux, Seedream, Recraft, Stable Diffusion, and Midjourney?
No. As of September 2026, Kunavo's image catalog includes Google's Nano Banana series, OpenAI's GPT-Image-2, and Ideogram v3 — 6 generation models and 3 editing models. Flux, Seedream, Recraft, Stable Diffusion, and Midjourney is not available on Kunavo. If your task specifically requires one of those, another provider is the right choice. Ideogram was added on September 14, 2026, and is billed by rendering speed rather than resolution, so compare prices at the speed you would actually use.
Which models can edit images?
As of September 2026, there are 3 editing models. Nano Banana Edit costs $0.0273 per image, GPT-Image-2 Edit costs $0.0633 per image, and Ideogram v3 Edit starts at $0.02625 per image depending on rendering speed. They support different editing workflows. Google's and OpenAI's editing models accept multiple source images in one request and take editing instructions only through the prompt — suitable for compositing product images into a background. Ideogram v3 Edit is the only model on this key that accepts masks (specifying that only the painted areas should change), making it suitable for retouching workflows.
Why are image APIs billed per image rather than per token?
Because cost depends on the fixed GPU work required per output, not on the input length. A five-word prompt and a 200-word prompt each create one image and take nearly the same rendering time, so token-based billing would charge for the wrong thing. In practice, this means image costs and text costs use different units and cannot be combined in one calculator.
Is there a free image generation API?
There is no free option suitable for production use. Some providers offer trial credits and consumer free tiers with daily limits, but each generation uses real GPU time, so no one offers an unlimited free image API. A realistic minimum is a few cents per image, and Kunavo's cheapest production rate is under 3 cents. Before using an API that advertises unlimited free generations, check what it actually routes requests to and how it stores prompts.