
Batch image generation API: queue, webhooks, safe retries
How to send many requests and keep track of every one: a queue on your side, webhooks, and the X-Idempotency-Key header for safe retries.
Эта страница есть на русском
Esta página también está en español
此页面也有中文版
#Blog category
Step-by-step examples in curl, Python and JavaScript: the request, the status, webhooks and safe retries.
4 articlesCategory RSS

How to send many requests and keep track of every one: a queue on your side, webhooks, and the X-Idempotency-Key header for safe retries.

When it is more convenient to use Google’s or OpenAI’s own API directly, and when a unified API suits you better: price, keys, limits, and payment. A list of pros and cons.

A key, one POST request, and a status check: a step-by-step example in curl, Python, and Node.js, with resolution up to 4K and a choice of aspect ratios.

The request, the status and the result for Nano Banana Pro, step by step in curl, Python, and Node.js, with the same scene at 1K, 2K, and 4K.

Nano Banana and GPT Image side by side: resolutions, price per image, average response time and what each model is best for.

Why a model sometimes returns no image, what it looks like in the API response, and how to reword a harmless prompt that was misread.

What a 429 response means, the limits behind it, and how to stay within them: how many requests to keep in flight, how to retry after a pause, and how webhooks replace frequent polling.