Retention Period and Storage Region of Input/Output Data in Gemini API Batch Mode (original) (raw)

December 17, 2025, 10:54am 1

I have questions regarding the retention period and storage region for input/output data in Gemini API’s Batch mode.

Currently, I am using the paid API model gemini-2.5-flash-preview-09-2025 to analyze PDFs and images via the Batch API.

I have confirmed that after a Batch job reaches BATCH_STATE_SUCCEEDED, I can retrieve the output results using batches.get. However, I cannot clearly determine from the documentation how long this output data is retained on Google’s side and until when it remains accessible.

Reference:

Therefore, I would appreciate clarification on the official specifications regarding the following points:

Retention Period and Storage Region for Input Data (Prompts/Files)

Prompts (text input):

• How long are prompts submitted to the Batch API retained on Google’s side?
• Also, in which region are they stored?

File Input (PDFs/Images):

• The documentation clearly states that files uploaded using the Files API have a retention period of 48 hours. However, when data is sent as inlineData, how long is this input data retained?
• Is it correct to understand that files are stored in the same region as prompts?

Retention Period and Storage Region for Output Data After BATCH_STATE_SUCCEEDED

• How long is the output data generated after a Batch job reaches BATCH_STATE_SUCCEEDED retained on Google’s side?
• In which region is it stored?

For operational purposes, I am planning automation for result retrieval and considering data management policies (storage and deletion design). It would be very helpful if you could provide information on the official retention period, storage region, and expected usage patterns.

Thank you in advance.