How is image generation billed?
Per image generated. The rate may vary by model, resolution or action, with current tiers shown on the pricing page. Failed generations are not charged.
Image generation API gateway
Generate, upscale, vary and edit images through a single gateway account. SuperAI API exposes GPT-Image-2, the full Midjourney action set, Grok Imagine and Google's Gemini image models with per-image billing and one set of credentials.
Recharge value
A $15 recharge adds gateway usage value benchmarked to $100 at official API list prices.
Equivalent usage depends on the model, tokens, resolution and selected group. This is SuperAI API balance, not an official provider account or transferable provider credit.
Recharge calculator
Enter a recharge amount to estimate the API usage value benchmarked against official provider list prices.
This estimates SuperAI API gateway usage value, not transferable credit in an official provider account. Actual consumption depends on the model, tokens, resolution and selected group.
curl -X POST https://superaiapi.com/v1/images/generations \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-image-2",
"prompt": "A cinematic city skyline at blue hour",
"size": "16:9"
}'The catalog mirrors what is actually live on the gateway, so what you evaluate is what you ship.
Image models are billed per generated image, with rates that can vary by model, resolution or action — no monthly seats and no idle subscription cost. Current tiers are published on the live pricing page, and usage logs show exactly what each generation cost.
GPT-Image-2 works with any OpenAI-compatible SDK by changing the Base URL and API key. Midjourney uses an asynchronous task flow with submit and fetch endpoints. Grok Imagine and Gemini image models are exposed through standard gateway endpoints with the same token.
Per image generated. The rate may vary by model, resolution or action, with current tiers shown on the pricing page. Failed generations are not charged.
Yes. GPT-Image-2 accepts the DALL·E 3 request format, so OpenAI-compatible SDKs work by changing the Base URL, API key and model name.
Yes. Imagine, upscale, variation, inpaint, zoom, pan, blend, describe and related actions are available through the async Midjourney-proxy API.