Quick answer
Gemini API documents batch/file storage as 2 GB per input file; 20 GB project file storage. The file and aggregate storage ceilings are independent of prompt context capacity.
Verified Aug 22, 2026Official source
Current limits
| Constraint | Current value | Scope | Verified source |
|---|---|---|---|
| batch/file storageThe file and aggregate storage ceilings are independent of prompt context capacity. | 2 GB per input file; 20 GB project file storage | Per project | GoogleAug 22, 2026 |
Why does this limit matter?
Large batch inputs can exhaust project storage even when individual files remain below 2 GB.
This value is scoped to Per project; a different plan, runtime, model, endpoint, region, or account can produce a different effective constraint.
What should you check?
- Inspect project file inventory, per-file bytes, and model context before scheduling work.
- Confirm the exact plan, model, runtime, endpoint, region, and account that serve the failing workload.
- Record the observed value, response headers or configuration, timestamp, and source without logging secrets.
Important caveats
- Uploaded bytes do not imply that a single model request can process the full file within its token window.
- Treat the official source and live account configuration as authoritative if they differ from this verified snapshot.
HyperObserve reports the documented platform constraint. Your application, SDK, gateway, provider, region, or account can impose a lower effective limit.
Related