# Gemini API Now Logs Your Requests by Default. Check This Setting > New Gemini API paid projects store requests by default for up to 55 days. How to switch logging off before it contradicts your privacy promises. - Canonical: https://zeroraised.com/blog/gemini-api-logging-on-by-default - Published: 2026-10-31 - Author: ZeroRaised Team (https://zeroraised.com/blog/authors/zeroraised) - Publisher: ZeroRaised (https://zeroraised.com/) - Section: Growth - Tags: artificial-intelligence, developer-tools, saas, bootstrapping The Gemini API now stores your prompts and responses by default for new paid projects, and keeps them for up to 55 days. Google made the change on 26 Oct 2026, so a project created on or after that date logs Generate Content API requests unless you set `store=false` or switch logging off in the dashboard ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). If your site, contract or privacy policy says you do not retain customer prompts, this is the setting that decides whether that claim is true. **TL;DR** - New paid Gemini API projects store Generate Content requests by default from 26 Oct 2026. - Existing projects keep their previous setting, so check rather than assume. - Logs are retained for a default maximum of 55 days, configurable to 7, 14, 28 or 55. - Logs are private to your Cloud project and are not used for product improvement unless you share datasets. - Set `store=false` per request, or turn off Generate Content API logging in the Logs and Datasets dashboard. **Disclosure:** ZeroRaised publishes this analysis and is a launch site for bootstrapped products. Facts come from the linked sources as checked on 31 Oct 2026; ZeroRaised has no commercial relationship with the companies named. **Default-on request logging** is an API setting where the provider stores each prompt and response unless the developer explicitly opts out, which makes retention the baseline rather than the exception. ## What changed in the Gemini API on 26 October 2026? New projects using the Generate Content API on the paid tier now store requests by default ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). The change covers `generateContent`, `streamGenerateContent` and `batchGenerateContent`, and brings them into line with the Interactions API, which already behaved this way ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). The detail that matters most for existing products is the carve-out: projects that already existed keep their current setting ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). So two founders can be on the same model and the same tier with different defaults, purely because of when they created the project. That is not a problem if you know which group you are in. It is a problem if you assume. ## How long does Google keep Gemini API logs? Logs are developer-owned API data for billing-enabled projects, private to your Google Cloud project, and separate from abuse-monitoring logs ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). Logging is on by default for Interactions API calls and for Generate Content API calls in projects created on or after 26 Oct 2026 ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). Retention has a default maximum of 55 days, and you can set it to 7, 14, 28 or 55 days ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). Datasets built from logs have no set retention period at all ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). That last line is easy to skim past. A log window you can reason about is not the same thing as a derived dataset that lives until someone deletes it. ## Does default logging mean Google trains on my prompts? No, not by default. Prompts and responses in logs are not used for product improvement unless you choose otherwise ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). The default is storage for operations, not a training feed. That changes the moment you share datasets with Google. Shared datasets are processed under the Unpaid Services terms, may be used to train models, and human reviewers may read them ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). This is a deliberate trade, not an accident, and it should be a documented decision inside your company rather than a checkbox someone clicks while exploring the console. The Contributor Tier is the formal version of that trade. It is an invite-only private preview that offers up to 50% off token costs on select models, including Gemini 4 Argon, Gemini 3.1 Pro Preview, Gemini 3.8 Flash and Gemini 3.5 Flash-Lite, in exchange for allowing Google to use data from those models under the Unpaid Services terms ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). Google says not to submit sensitive or personal information ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-policy)). ## How do I turn Gemini API logging off? There are two levers. The precise one is `store=false` on individual requests ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). Use it on the routes that carry customer content you promised not to keep, such as a document summariser or a support assistant. The blunt one is turning off Generate Content API logging in the Settings panel of the Logs and Datasets dashboard ([Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog)). That applies at the project level and is the right switch when your default should be "do not store" and exceptions should be rare. The dashboard is where you manage logs and datasets ([Gemini API docs](https://ai.google.dev/gemini-api/docs/logs-datasets)). Pick one owner for that panel. One pattern: keep a separate Cloud project for debugging, so an experiment cannot quietly change the retention posture of production. ## What should your privacy policy say? Write down four things plainly. First, which vendor and which tier you use, because the Gemini terms differ sharply between paid and unpaid use. Second, whether prompts are retained, and for how long. Third, whether they are used to improve or train models, and whether you share datasets. Fourth, who can read them. Then check that the answer matches the console. A policy is a description of your configuration, not a wish. The [security checklist for vibe-coded apps](https://zeroraised.com/blog/vibe-coded-app-security-checklist) is a useful prompt to review what you send. If you use several AI vendors, the [guide to choosing an LLM API](https://zeroraised.com/blog/how-to-choose-an-llm-api) is a useful way to compare data terms before you add another line to that list. If you are weighing retention against cost, the [LLM cost guide](https://zeroraised.com/blog/how-to-reduce-llm-costs) covers the trade-offs. Do not let a discount programme decide your data posture by default. ## What this means for bootstrapped founders - **Check your project age.** New projects log by default; older projects keep their setting. Confirm which one you have instead of guessing. - **Inventory what you send.** List the routes that carry customer content, then decide which ones need `store=false`. - **Keep the default boring.** If most of your traffic is sensitive, turn logging off at the project level and opt in for debugging. - **Never promise more than the setting delivers.** "We do not train on your data" and "we do not store your data" are different claims. - **Treat data discounts as a product decision.** A cheaper token price in exchange for data sharing belongs in the same conversation as your terms, not only your budget. The wider comparison of vendor defaults is in [do AI APIs train on your data](https://zeroraised.com/blog/do-ai-apis-train-on-your-data). The short version: the tier and the setting matter more than the brand on the invoice. ## Where ZeroRaised fits ZeroRaised is a launch and discovery site for products built without institutional venture capital or accelerator equity. If you have a product people can try, you can [submit your product](https://zeroraised.com/submit). The [pricing page](https://zeroraised.com/pricing) explains the current options, and there are no promises of traffic or results. If this kind of plain analysis is useful, [join the ZeroRaised newsletter](https://zeroraised.com/newsletter). It is a good place to hear when vendor defaults change, because defaults are where quiet surprises live. ## Sources - [Gemini API release notes, 26 Oct 2026](https://ai.google.dev/gemini-api/docs/changelog) - [Gemini API: Data logging and sharing](https://ai.google.dev/gemini-api/docs/logs-policy) - [Gemini API: Logs and datasets](https://ai.google.dev/gemini-api/docs/logs-datasets) ## Frequently asked questions ### Is Gemini API logging on by default? For new projects on the paid tier that use the Generate Content API, yes. Requests are stored by default from 26 Oct 2026, matching the Interactions API. Projects created before that date keep the setting they already had, so the answer depends on when the project was created. ### How long does Google retain Gemini API logs? The default maximum is 55 days. You can set retention to 7, 14, 28 or 55 days. Datasets you build from logs have no set retention period, which is a separate decision from the log window. ### Does Gemini use my logged prompts to train models? By default, prompts and responses in logs are not used for product improvement. That changes if you choose to share datasets with Google, which are processed under the Unpaid Services terms, may be used to train models, and may be read by human reviewers. ### How do I turn off Gemini API logging? Set store=false on individual requests, or turn off Generate Content API logging in the Settings panel of the Logs and Datasets dashboard. The first option is surgical; the second applies to the project. ### Does this affect the free tier? The change is described for new paid projects using the Generate Content API. The free tier has its own data terms, and those terms deserve a separate read before you send customer content through it.