Checked on October 26, 2026

OpenRouter alternatives for image generation

OpenRouter is a gateway with one key and one balance for more than 500 models and 80 providers. Images there are counted in tokens, so instead of a price per image the table shows per-token rates and an explanation of how to read them.

Compare prices Get started

A Rendergrid article. We’re one of the parties in this comparison, so the OpenRouter figures come only from its public pages, each with a link and a date, and below we say separately when OpenRouter is the better choice. How we check and correct

01Who it is

What OpenRouter is

Briefly, and only from their own pages. Website: openrouter.ai.

OpenRouter (openrouter.ai) is a gateway to models from different providers: the pricing page states more than 500 models and more than 80 providers. By OpenRouter’s own account, it doesn’t mark up inference: a model’s price equals the provider’s price, and it takes its own fee (5.5% on Standard, 8% on Business) when you buy credits.

Images are generated by a separate Image API: POST /api/v1/images returns the image in base64 right in the response (some models can stream previews). For our models, OpenRouter’s page names a rate per output-image token; the total is reported after the request by the usage.cost field, so the price of an image can’t be calculated in advance.

02Pricing

Prices for our models

How much one image costs at OpenRouter and at Rendergrid. The OpenRouter prices are from its public pages, checked on October 26, 2026.

OpenRouter has no per-image prices on the checked pages that can be compared with ours by resolution.

Image prices: OpenRouter and Rendergrid
Model	Resolution	OpenRouter	Rendergrid	Difference
Nano Banana Pro`nano-banana-pro`	1K	by tokens$120 per 1M tokens¹	$0.04	can’t be compared
Nano Banana Pro	2K	by tokens$120 per 1M tokens	$0.04	can’t be compared
Nano Banana Pro	4K	by tokens$120 per 1M tokens	$0.07	can’t be compared
Nano Banana 2`nano-banana-2`	1K	by tokens$60 per 1M tokens²	$0.025	can’t be compared
Nano Banana 2	2K	by tokens$60 per 1M tokens	$0.04	can’t be compared
Nano Banana 2	4K	by tokens$60 per 1M tokens	$0.06	can’t be compared
Nano Banana 2.1`nano-banana-2.1`	1K	by tokens³	$0.02	can’t be compared
Nano Banana 2.1	2K	by tokens	$0.03	can’t be compared
Nano Banana 2.1	4K	by tokens	$0.04	can’t be compared
GPT Image 2`gpt-image-2`	1K	by tokens$30 per 1M tokens⁴	$0.02	can’t be compared
GPT Image 2.5`gpt-image-2.5`	1K	not availablenot on their pages⁵	$0.025	can’t be compared

1Nano Banana Pro. Rate: $120 per 1M tokens (output image; input—image $2 per 1M tokens). The rate is that of the providers Google AI Studio and Google Vertex; 1K and 2K are available on Vertex. Source: openrouter.ai/google/gemini-3-pro-image. Checked on October 26, 2026.
2Nano Banana 2. Rate: $60 per 1M tokens (output image). The rate is that of the providers Google AI Studio and Google Vertex; sizes 512, 1K, 2K, and 4K. Source: openrouter.ai/google/gemini-3.1-flash-image. Checked on October 26, 2026.
3Nano Banana 2.1. Provider: Google AI Studio; sizes 1K, 2K, and 4K. Source: openrouter.ai/google/gemini-nano-banana-2.1. Checked on October 26, 2026.
4GPT Image 2. Rate: $30 per 1M tokens (output image; input—text $5, image $8 per 1M tokens). Example on the OpenRouter page: quality high, 16:9, 1536×864—$0.13, 94 seconds. Source: openrouter.ai/openai/gpt-image-2. Checked on October 26, 2026.
5GPT Image 2.5. We didn’t find a “GPT Image 2.5” page without a suffix; there are GPT Image 2.5 Flare and GPT Image 2.5 Sunburst, in the rows below. Source: openrouter.ai/openai/gpt-image-2.5-flare. Checked on October 26, 2026.

The difference is measured from the larger of the two prices, rounded down to a whole percent. The “≈” sign is an estimate that the provider’s own page gives. For GPT Image, quality and size differ between providers: each one’s conditions are in the footnote. We compare prices, not image quality. Nano Banana Flash, Nano Banana 2 Lite, and GPT Image 2 HD are not in the table: we haven’t checked the provider’s prices for them. Our own prices and the official Google and OpenAI prices are on the “Pricing” page.

03Terms

Payment, starter credits, and operating mode

What OpenRouter’s pages say, next to our terms. Where a page is silent, we say so.

Terms: OpenRouter and Rendergrid

Condition	OpenRouter	Rendergrid
Payment	Cards, Alipay, and cryptocurrency (USDC); the fee on buying credits is 5.5% on Standard and 8% on Business; there is no minimum payment and no commitment; invoices for Enterprisesource	USDT (crypto top-up): the full amount sent is credited to the balance
Credits at no charge	The Free plan: 25+ free models with a limit of 20 requests per minute and 50 per day (after buying $10 of credits, 1,000 per day). We found no trial balance for paid models on the checked pagessource	There’s no trial balance
Synchronous or asynchronous	Synchronous: POST /api/v1/images returns the image in base64 in the response itself; some models can stream previews (SSE). The checked pages have no polling or deferred taskssource	Asynchronous: POST returns 202 and an ID; the result comes through polling GET /creations/{id} (no more than once every 5 seconds) or a webhook
Webhooks	not found: the Image API documentation has no webhooks: the result comes in the response to the request itself	Yes: webhook_url in every request; the creation.completed and creation.failed events; an HMAC-SHA256 signature over the timestamp and the whole request body
Failed request	A failed or canceled generation isn’t charged (a 502 response). If the client disconnects, the provider may still finish the image, but OpenRouter doesn’t bill you for what you didn’t receivesource	If a generation fails, the charge is returned to the balance automatically; a declined request is refunded the same way
Limits	There are no plan limits on paid models: the restrictions come from the providers, and OpenRouter answers 429 only on very sharp spikessource	By default for a new account: 120 requests per minute and up to 100 generations queued or running; account limits can be adjusted to match your traffic
Result storage	not checked: the image comes in the response itself; the storage period is not named on the checked pages	Result files are stored for one day

Also on their pages

You can plug in your own provider key (BYOK): Standard and Business include $25,000 a month without OpenRouter’s fee, then 5%. source
Routing between providers of the same model: order, an allowed list, sorting by price, speed, or latency, and disabling fallbacks. source
Regional routing (EU and US) is included in the Business and Enterprise plans. source

Our terms in detail are in the documentation.

04Strengths

Where we’re stronger, and where OpenRouter is

Only what can be verified from public pages and our documentation.

Where Rendergrid is stronger

For our models, the price of one image is stated by resolution. At OpenRouter, the model page gives a per-token rate, and the amount for an image is reported only in the response (the usage.cost field).
The request is asynchronous: you get an ID right away and don’t hold a connection while the image is being prepared (in the example on OpenRouter’s page, it took 94 seconds). The result comes through a signed webhook or by polling. At OpenRouter the request is synchronous, and the checked pages have no webhooks for images.
Limits are stated as numbers: by default 120 requests per minute and up to 100 generations queued or running. At OpenRouter, the limits on paid models are set by the providers.

Where OpenRouter is stronger

One key and one balance for 500+ models and 80+ providers, including language models.
Routing between providers of the same model: you can set the order, restrict the list, and turn off the fallback route.
Payment by card, Alipay, and USDC; you can plug in your own provider key (BYOK).
A free plan with 25+ free models.
No plan limits on paid models: the restrictions come from the providers.
The image arrives in the response itself, with no polling or webhook; some models can stream previews.

05Choosing

When to pick OpenRouter instead

We write this page as a participant in the comparison, so we name separately the cases where OpenRouter suits you better.

Pick OpenRouter if

You need text, video, and other models on one key together with images.
You need routing between providers of the same model and a manual choice of order.
You pay by card, Alipay, or USDC, and token-based billing works for you.
You need a simple synchronous call: the image in the response itself, with no polling or webhook.
You want to plug in your own provider key and pay the provider directly.

06Migration

Migrating to Rendergrid in three steps

If you decide to give it a try, here is what to change in the code. Parameters and responses are in the documentation.

Key and address
At OpenRouter, requests go to https://openrouter.ai/api/v1/images with an Authorization: Bearer header. For us it is https://api.rendergrid.io/api/public/v1 with the same Authorization: Bearer header (or X-API-Key).
Request
Instead of POST /api/v1/images, call POST /images/generate: model, prompt, aspect_ratio, and resolution sit at the root of the body, but the model IDs are ours (for example, nano-banana-pro, not google/gemini-3-pro-image). The provider.* fields (provider selection) are not needed in our request. The webhook is set with the webhook_url field. Add X-Idempotency-Key.
Result
Instead of a base64 image in the response body, you’ll get a 202 response with an ID: poll GET /creations/{id} (queued, processing, completed, and failed; links in result_urls) or accept the creation.completed webhook. Our files are kept for a day, so download them right away.

POST /images/generate · example request

IDEMPOTENCY_KEY="$(uuidgen)"
curl -X POST https://api.rendergrid.io/api/public/v1/images/generate \
  -H "Authorization: Bearer $RENDERGRID_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Idempotency-Key: $IDEMPOTENCY_KEY" \
  -d '{
    "model": "nano-banana-2",
    "prompt": "a ceramic mug on a marble table",
    "resolution": "2K",
    "aspect_ratio": "1:1"
  }'

07FAQ

Frequently asked questions

How do you pay at OpenRouter and at Rendergrid?

Cards, Alipay, and cryptocurrency (USDC); the fee on buying credits is 5.5% on Standard and 8% on Business; there is no minimum payment and no commitment; invoices for Enterprise. At Rendergrid: USDT (crypto top-up): the full amount sent is credited to the balance.

Does OpenRouter give credits at no charge?

The Free plan: 25+ free models with a limit of 20 requests per minute and 50 per day (after buying $10 of credits, 1,000 per day). We found no trial balance for paid models on the checked pages. At Rendergrid there’s no trial balance.

Does OpenRouter work synchronously or asynchronously?

Synchronously: POST /api/v1/images returns the image in base64 in the response itself; some models can stream previews (SSE). The checked pages have no polling or deferred tasks. Rendergrid is asynchronous: POST returns 202 and an ID; the result comes through polling GET /creations/{id} (no more than once every 5 seconds) or a webhook.

How do I move from OpenRouter to Rendergrid?

Three steps: 1) key and address; 2) request; 3) result. The details and an example request are in the “Migration in three steps” section on this page.

Why does the table have no price per image for OpenRouter?

On OpenRouter’s pages, Nano Banana and GPT Image have rates per output-image token, and the total depends on the request: it’s reported by the usage.cost field in the response. So we don’t invent a price per image; we show the rates. Rendergrid’s prices are stated per image and resolution: they’re in the table above.

Where do these figures come from, and when were they checked?

All the OpenRouter data comes from its public pages, with links in the “Sources” section. Checked on October 26, 2026. Rendergrid wrote this page, so we give a source for every figure.

08Sources

Sources and check date

All the OpenRouter data was taken from these pages, checked on October 26, 2026. Prices change: open the provider’s page before you decide.

1
Nano Banana Pro on OpenRouter: provider rates (their public API)openrouter.ai/api/v1/images/models/google/gemini-3-pro-imag… · checked on October 26, 2026
2
Nano Banana 2 on OpenRouter: provider ratesopenrouter.ai/api/v1/images/models/google/gemini-3.1-flash-… · checked on October 26, 2026
3
Nano Banana 2.1 on OpenRouter: provider ratesopenrouter.ai/api/v1/images/models/google/gemini-nano-banan… · checked on October 26, 2026
4
GPT Image 2 on OpenRouter: provider ratesopenrouter.ai/api/v1/images/models/openai/gpt-image-2/endpo… · checked on October 26, 2026
5
GPT Image 2.5 Flare on OpenRouter: provider ratesopenrouter.ai/api/v1/images/models/openai/gpt-image-2.5-fla… · checked on October 26, 2026
6
GPT Image 2.5 Sunburst on OpenRouter: provider ratesopenrouter.ai/api/v1/images/models/openai/gpt-image-2.5-sun… · checked on October 26, 2026
7
OpenRouter: image generation (Image API)openrouter.ai/docs/guides/overview/multimodal/image-generat… · checked on October 26, 2026
8
OpenRouter: pricing and feesopenrouter.ai/pricing · checked on October 26, 2026
9
OpenRouter: limitsopenrouter.ai/docs/api_reference/limits · checked on October 26, 2026
10
OpenRouter: FAQopenrouter.ai/docs/faq · checked on October 26, 2026
11
OpenRouter: home pageopenrouter.ai · checked on October 26, 2026

Found an inaccuracy? Write to support@rendergrid.io. The correction process is described in “About the editors”.

Try Rendergrid

One price per image and resolution, signed webhooks, an automatic refund for a failed request.

Get started All comparisons