Data retention - verified
Does the Gemini API store your API prompts?
Default retention: 55 days, abuse monitoring (paid tier). Trains on your API data: Depends on tier (Paid tier: No. Unpaid tier: Yes, except in the EEA, Switzerland and the UK). Zero data retention: not offered (no exemption from abuse logging documented; Google points to Vertex AI). Regional processing: none documented.
Scope: Gemini API and AI Studio. Not the Gemini consumer app. Paid vs unpaid tiers have different terms.
Read from ai.google.dev. Compare every AI API's data retention
- Trains on data
- Depends on tier
- ZDR
- Not offered
- Regions
- None
- Verified
- 2026-10-18
How long are prompts kept by default?
- Default retention
55 days, abuse monitoring (paid tier)
Paid Services: prompts and responses logged for a limited period solely for abuse detection; the abuse monitoring page sets this at 55 days. Unpaid Services: content used to improve products, with human review possible.
“Google retains the following data for fifty-five (55) days for the purposes of detecting and preventing violations of the Prohibited Use Policy to maintain the safety and security of the Services, and any required legal or regulatory disclosures:”
ai.google.dev “For Paid Services, Google logs prompts and responses for a limited period of time, solely for detecting and preventing violations of the Prohibited Use Policy to maintain the safety and security of the Services, and any required legal or regulatory disclosures.”
ai.google.dev - Duration
55 days (abuse monitoring)
- Purpose
detecting and preventing violations of the Prohibited Use Policy, safety and security, legal or regulatory disclosures
Does it train on your API data?
Depends on tier
Paid tier: No. Unpaid tier: Yes, except in the EEA, Switzerland and the UK
Recorded as: Paid: no. Unpaid (free tier / AI Studio without billing): yes, except users in the EEA, Switzerland or UK, where paid-service terms apply to all services.
“When you use Paid Services, including, for example, the paid quota of the Gemini API, Google doesn't use your prompts (including associated system instructions, cached content, and files such as images, videos, or documents) or responses to improve our products”
“When you use Unpaid Services, including, for example, Google AI Studio and the unpaid quota on Gemini API, Google uses the content you submit to the Services and any generated responses to provide, improve, and develop Google products and services and machine learning technologies, including Google's enterprise features, products, and services, consistent with our Privacy Policy.”
“If you're in the European Economic Area, Switzerland, or the United Kingdom, the terms under "How Google uses Your Data" in "Paid Services" apply to all Services, including Google AI Studio and unpaid quota in the Gemini API, even though they are offered free of charge.”
Zero data retention (ZDR)
- Availability
Not offered
No exemption from abuse logging documented; Google points to Vertex AI
Recorded as: Partial / not guaranteed: no ZDR exemption from abuse-monitoring prompt logging is documented for the Gemini Developer API; Google points to Vertex AI (Gemini Enterprise Agent Platform) for guaranteed ZDR. Other storage can be avoided by settings (store=false, no session resumption, no explicit caching, deleting files, avoiding Maps grounding).
- Who can get it
Not stated as an approval program for the Gemini Developer API.
- Not covered
Grounding with Google Maps (30-day storage cannot be disabled); Interactions API stores state by default; Live API session resumption (up to 24h); Files API; explicit context caching.
- From the docs
“If your workload requires guaranteed zero data retention or enterprise data processing agreements, use Vertex AI.”
ai.google.dev “Customer data is typically retained for limited periods of time in the following scenarios and conditions. To achieve zero data retention, customers must take specific actions or avoid specific features within each of these areas:”
ai.google.dev “By default, Gemini models cache data in-memory to reduce latency and cost for developers. This data is strictly in RAM (not at-rest), isolated at the project level, and has a 24-hour TTL. This does not violate zero data retention.”
ai.google.dev
Features kept longer
Stateful features hold your content until a job finishes, a window expires or you delete it, often regardless of the default above.
| Feature | Retention | The docs say |
|---|---|---|
| Interactions API (stored by default, store=true) | Paid tier 55 days (configurable to 7, 14, 28 or 55 days in AI Studio); free tier 1 day | “By default, the API stores all Interaction objects (store=true) in order to simplify use of server-side state management features (with previous_interaction_id), background execution (using background=true) and observability purposes.”ai.google.dev |
| Interactions API retention window | Paid 55 days; free 1 day | “Paid tier : The system retains interactions for 55 days . Free tier : The system retains interactions for 1 day .”ai.google.dev |
| Files API | 48 hours | “Files are stored for 48 hours.”ai.google.dev |
| Batch API results | 6 weeks | “By default, batch job results are stored and available for download for 6 weeks, before they are permanently deleted.”ai.google.dev |
| Grounding with Google Maps | 30 days (cannot be disabled) | “You acknowledge that it is reasonably necessary for Google to store prompts, contextual information that you may provide, and generated content for thirty (30) days for the purposes of creating Google Maps Grounded Results”ai.google.dev |
| Live API session resumption | up to 24 hours | “If a session handle is generated, conversation state (including text, audio, and video) is retained for up to 24 hours.”ai.google.dev |
| Explicit context caching (cached_content) | User-defined ttl or expire_time | “While the logs of these requests follow ZDR dropping policies, the cached context itself is stored with a user-defined ttl or expire_time.”ai.google.dev |
| Implicit in-memory caching | 24-hour TTL, RAM only | “This data is strictly in RAM (not at-rest), isolated at the project level, and has a 24-hour TTL.”ai.google.dev |
| Model tuning content | Retained with tuned model; deleted when tuned model is deleted | “When you delete a tuned model, the related tuning content is also deleted.”ai.google.dev |
Where is your data stored and processed?
- Regional processing
None documented
- Stated location
Logged data may be stored in any country where Google or its agents have facilities
- How to choose
- Not stated
- What it covers
No data residency control is documented for the Gemini Developer API. Terms state logged data may be stored transiently or cached in any country where Google or its agents maintain facilities.
- From the docs
“This data may be stored transiently or cached in any country in which Google or its agents maintain facilities.”
ai.google.dev “You may use only Paid Services when making API Clients available to users in the European Economic Area, Switzerland, or the United Kingdom.”
ai.google.dev
Regional endpoints
Regional hostnames: Not stated
Certifications
Certifications on the pages checked: Not stated
Research notes
- https://ai.google.dev/gemini-api/docs/data-residency returned 404 on 2026-10-18; no residency page found.
- Gemini API access is a Paid Service only through a Cloud Project with an active billing account.
- Paid Services processing follows the Data Processing Addendum for Products Where Google is a Data Processor.
- Abuse-monitoring data is not used to train models other than those used for policy enforcement (usage-policies page).
FAQ
Does the Gemini API store your API prompts?
55 days, abuse monitoring (paid tier). Paid Services: prompts and responses logged for a limited period solely for abuse detection; the abuse monitoring page sets this at 55 days. Unpaid Services: content used to improve products, with human review possible. The provider says: "Google retains the following data for fifty-five (55) days for the purposes of detecting and preventing violations of the Prohibited Use Policy to maintain the safety and security of the Services, and any required legal or regulatory disclosures:"
Does the Gemini API train on your API data?
Depends on tier (Paid tier: No. Unpaid tier: Yes, except in the EEA, Switzerland and the UK). The provider says: "When you use Paid Services, including, for example, the paid quota of the Gemini API, Google doesn't use your prompts (including associated system instructions, cached content, and files such as images, videos, or documents) or responses to improve our products"
Does the Gemini API offer zero data retention?
Not offered (no exemption from abuse logging documented; Google points to Vertex AI). Not stated as an approval program for the Gemini Developer API.
Where does the Gemini API process your data?
Regional processing options: None documented. Stated processing location: Logged data may be stored in any country where Google or its agents have facilities.
Sources
Last verified from Gemini API's own pages below. Quotes are verbatim.
This page summarizes what each provider's own pages say. It is not legal advice; your contract, data processing addendum and the provider's current terms govern what happens to your data.
Newsletter
Know when a provider changes its data terms
One email per week. New model releases, pricing and policy changes, and what they mean for what you build.