Text-to-image and image-to-image through one endpoint — GPT Image, FLUX, Seedream, Nano Banana, Qwen-Image and more — with per-image prices compared live.
black-forest-labs/flux.2-dev2 hosts · per image
google/nano-banana-22 hosts · per image
google/nano-banana-pro2 hosts · per image
black-forest-labs/flux.2-pro3 hosts · per image
black-forest-labs/FLUX.1-dev2 hosts · per image
seedream-5.0-flash2 hosts · per image
alibaba/wan-2.62 hosts · per image
black-forest-labs/flux.2-max3 hosts · per image
seedream-5.0-pro3 hosts · per image
black-forest-labs/flux.2-flex1 host · per image
recraft-4.12 hosts · per image
bytedance/seedream-4.04 hosts · per image
bytedance/seedream-5.0-lite4 hosts · per image
qwen-image-max2 hosts · per image
grok-imagine-image-quality1 host · per image
grok-imagine-image2 hosts · per image
grok-imagine-image-2.01 host · per image
prunaai/p-image2 hosts · per image
black-forest-labs/flux-1.1-pro4 hosts · per image
black-forest-labs/flux.2-klein-4b2 hosts · per image
black-forest-labs/flux-2-klein-9b2 hosts · per image
black-forest-labs/flux-kontext-pro2 hosts · per image
black-forest-labs/flux-kontext-max2 hosts · per image
black-forest-labs/flux-3-image1 host · per image
openai/gpt-image-1.52 hosts · per image
ideogram-ai/ideogram-4.02 hosts · per image
ideogram-ai/ideogram-4.0-quality2 hosts · per image
cosmos3-super-text2image2 hosts · per image
muse-image3 hosts · per image
krea-2-medium-turbo2 hosts · per image
openai/gpt-image-24 hosts · per image
qwen-image-edit2 hosts · per image
alibaba/z-image-turbo2 hosts · per image
seedream-4.51 host · per image
qwen-image-2.12 hosts · per image
curl https://videorouter.sh/api/v1/images \
-H "Authorization: Bearer $VIDEOROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "openai/gpt-image-1/openai", "prompt": "a red panda astronaut floating in space", "aspect_ratio": "16:9"}'import requests
r = requests.post("https://videorouter.sh/api/v1/images",
headers={"Authorization": "Bearer llmr_sk_live_..."},
json={"model": "openai/gpt-image-1/openai", "prompt": "a red panda astronaut floating in space"}).json()
print(r["usage"]["cost"])Quickstart · image-to-image uses input_references on the same endpoint.
How to choose an image generation model and host: per-image price, quality tiers, editing support, and why you should compare the same model across providers.
Text-to-image versus image-to-image in an image generation API: how input_references works, editing workflows, which models support it, and common pitfalls.
A decision framework for choosing between GPT Image, FLUX, Seedream, Nano Banana and Qwen-Image on an image API, and how to A/B them on your own prompts.
How per-image billing works on an image generation API and the levers that cut spend: resolution, quality, draft-then-final, batching, caching and spend caps.
A parameter-by-parameter guide to calling GPT Image models through /v1/images: size controls, OpenAI-only options, edits, response format and cost levers.
The FLUX variants in one image API (dev, pro, max, flex, klein, FLUX 3 Image, Kontext): what the catalog says about each and how to A/B them.
The Seedream models in the catalog (4.0, 4.5, 5.0 Lite, Flash and Pro): how to find the right id, how to choose between versions and a harness to test them.
Call google/nano-banana-2 and nano-banana-pro with curl or Python: request shape, input_references for edits, response handling, hosts and common mistakes.
It turns a text prompt (optionally with reference images) into one or more generated images over HTTP.
Per image, usually varying by resolution or quality setting. The pricing page shows the cheapest host for each model.
Yes — pass the image in input_references with an instruction prompt.
It depends on style and budget. Compare outputs on the same prompt using the same endpoint, swapping only the model id.
VideoRouter puts it next to dozens of other video and image models behind one API key, so you can compare providers, prices and fail over automatically. Compare providers on VideoRouter →