Non-batch background generation

Gemini API
,
1

OpenAI models support a “background” option which lets you trigger a request, then poll for a response. Even better, they also support web hooks to notify you when a response is ready. This is especially useful for image generation, or other long-running tasks where a request could be interrupted or you could hit a wall-time limit on cloud workers.

I know Gemini has a Batch API, but that’s really designed for bulk off-line workflows rather than in-line flows. Is there a way to achieve this with Gemini?

3

Hi @Mark_K , Thanks for reaching out to us.

The scope of supported features aligns precisely with what is outlined in our official documentation. Please refer to our documentation on Batch API for comprehensive details.

4

Thanks Sonali. So the answer is “no, Gemini does not support this”