
Archive approved Imagild API outputs by downloading the authenticated result, validating it, and copying it into storage your application controls. Keep a manifest that links the task to the stored object and checksum. A private provider content route is not a permanent public archive, and an archival retry should not submit another generation.
Key Takeaways
- Complete generation, verified delivery, and verified archival storage are separate stages.
- Use a stable archive record so retries find the existing copy instead of multiplying assets.
- Apply your own access, retention, and deletion policy to customer media.
Establish what needs to be preserved
Record the task ID, owning account reference, request model and tier, approval decision, content type, byte count, and measured media properties. Keep the original operation identity so the archive can be traced back to the approved paid request.
The API console requires authentication and active membership for result delivery. Do not assume a successful generation grants an anonymous or lifetime download link.
Replicate's output-file guidance illustrates why generated-file delivery and retention deserve explicit handling. Its expiration durations do not define Imagild's retention contract.
Make archival work a recoverable stage
A useful application pipeline is:
- Confirm the saved task succeeded.
- Download its result through the authenticated content route.
- Validate type, completeness, and decoding.
- Compute an integrity checksum.
- Upload the approved bytes to your controlled destination.
- Verify the stored copy before marking the archive complete.
Use a stable archive identity based on the authorized account, task, and output version or checksum. Object keys should come from controlled identifiers, not an untrusted filename or prompt. Keep the bucket private unless deliberate publication has been approved.
If the worker restarts after upload but before saving completion, inspect the expected stored object and checksum. Do not create another generation merely because your archive status row is incomplete.
Compute a checksum without loading the full file
This native Node.js example computes SHA-256 over an already validated local file using a stream:
import { createReadStream } from "node:fs";
import { createHash } from "node:crypto";
async function sha256File(path) {
const hash = createHash("sha256");
for await (const chunk of createReadStream(path)) hash.update(chunk);
return hash.digest("hex");
}
console.log(await sha256File("approved-result.bin"));
It does not upload anything or establish that a remote object is correct. Compare the checksum with a verified stored representation through your storage provider's supported integrity procedure.
Amazon S3's integrity documentation describes object checksums for upload and download verification. S3 is one possible destination, not an Imagild feature or a requirement to use that provider.
Keep the manifest useful for recovery
| Manifest field | Recovery purpose |
|---|---|
| Task ID and owning account reference | Identifies the authorized source |
| Archive object key and storage location | Finds the controlled copy |
| SHA-256 and byte count | Compares integrity across transfers |
| Content type and decoded dimensions/duration | Confirms media interpretation |
| Approval and verification timestamps | Separates review from copy completion |
| Archive state and last error | Allows stage-specific retries |
Do not place the API key, signed download tokens, or unnecessary private prompts in a public manifest. An internal manifest may itself contain sensitive customer associations and needs access controls.
Choose a retention policy that fits rights, privacy, and operational requirements. Do not invent a universal provider retention period from an older website article or another service's documentation.
Test restoration before relying on the archive
Retrieve one approved stored object through your authorized application path and compare its checksum and media properties. A successful upload response alone is not a demonstrated restore.
The asset archive checklist provides general backup steps. Use the HTML-download diagnostic if the saved source is actually an error page.
Current generation choices and points remain in the model market. Your storage and delivery charges are separate application costs. A verified archive makes an existing approved result reusable without implying permanent provider URLs or authorizing another paid render.
Related tasks
- How to Archive Imagild Assets Before the 90-Day Storage Window Ends
- Why Does My Downloaded Image Open as a Web Page?
Sources and scope
- Imagild API console: Private authenticated result delivery and membership requirements.
- Imagild model market: New generation cost is distinct from application's archive storage costs.
- Replicate output files: Output delivery and retention require explicit application handling; provider duration not transferred.
- Amazon S3 object integrity: Storage checksum verification is a separate integrity control for uploaded/downloaded objects.
Product instructions were checked against the current Imagild API contract and the public model catalog. Code, budgets and acceptance checks are illustrative; this article does not claim a new paid generation test, an uptime measurement, a search-volume study or a ranking of AI citations.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: Original Imagild product screenshot captured on 28 October 2026; model availability and prices may change. This image is not an API response or a performance benchmark.