Short answer
Google’s Gemini app help page does not specify a fixed number of images you can generate. It describes general, compute-based limits for Gemini Apps, but does not give a separate image-generation allowance. So an exact app image count—and an image-specific reset schedule—cannot be confirmed from that guidance.
For the Gemini API, there is no single image count that applies to every developer. Limits depend on the model and project. Google documents an images-per-minute (IPM) rate limit for image-capable models; that API quota is not the same as an app allowance.
Gemini app: what Google’s limit guidance says
Google describes Gemini Apps usage limits as compute-based, factoring in prompt complexity, the models and features used, and chat length. Its help page says the general limit refreshes every five hours until a weekly limit is reached. That is general app-limit guidance, not a promise that a set number of image generations resets every five hours. The page also notes that limits may change.
Google lists plan limits as relative usage multipliers: AI Plus has 2× standard limits, AI Pro 4× standard limits, and AI Ultra 5× or 20× AI Pro limits depending on the subscription. These are not image counts, and the page reviewed does not specify how many images those multipliers correspond to.
Production Gemini API: check the model and project quota
Do not use a Gemini app allowance to plan API throughput. Gemini API rate limits are applied per project, not per API key; they vary by model and can include requests per minute, requests per day, and images per minute for image-capable models. Google says daily request quotas reset at midnight Pacific time.
For example, Google’s model documentation identifies gemini-3.1-flash-image as an image-generation model and lists a 32,768-token output limit. That is a token limit, not a stated number of images per request. The official model and image-generation materials reviewed here do not establish a model-specific per-request image count or an output-count setting for gemini-3.1-flash-image. IPM and the output-token ceiling alone cannot tell you how many images a single request returns. Check the active project and model limits in Google AI Studio before estimating capacity. (ai.google.dev)
If your active rate limits are insufficient, Google’s rate-limits documentation provides a request-a-rate-limit-increase path and describes usage tiers. An increase is not an image-count guarantee; confirm the resulting limit for your project and model in AI Studio.
If image generation stops
- In the Gemini app: Check the account and plan in use and read the limit notice. Google’s general five-hour refresh guidance is not an image-specific reset promise.
- In the API: Check the active rate limits for the correct project and model in AI Studio. Requests per day reset at midnight Pacific time, but that should not be assumed to describe the reset behavior of every other quota dimension.
Bottom line: Google’s reviewed Gemini app guidance does not establish a fixed image count. For API use, check the selected model’s live project quota; IPM measures rate, not a universal total number of images.
Sources
Gemini Apps limits & feature availability — Google Help