Upload formats and size limits: what works on the farm

Before you upload a project to Super Renders Farm, it helps to know which transfer method fits your file size, why the farm needs your project unpacked, and what happens to your files once a job finishes.

Quick answer

| Method | Ideal for | Limits |
|---|---|---|
| Web upload | Small projects (under ~2 GB), one-off submissions | No hard limit, but slow above ~2 GB and unreliable above ~5 GB, with no resume; above ~2 GB use the Client App or S3 access |
| Client App | Large projects, hands-off uploads and downloads | No practical limit — resumable, parallel, plus auto-download for output |
| S3 access (Cyberduck / AWS CLI) | Scripted or client-based transfers to your SRF Space | No limit set by us; upload the project folder unpacked, as with every path |

Archives: upload your project unpacked

The farm does not extract archives (.zip, .rar, .7z, .tar, .tar.gz), so nothing inside an archive gets rendered. Upload your project folder unpacked, with the scene file and every referenced asset in place.

Here's the practical effect: the web uploader accepts an archive file, and for .zip, .rar and .7z it warns that the file won't be unpacked automatically. The job still can't see the scene inside it. If you received a project as an archive, extract it on your workstation first, then upload the extracted folder.

Every upload path takes a plain folder: drag it into the web uploader, upload it from the Client App's Spaces tab, or copy it to your SRF Space over S3 access.

Uploading large projects (above ~2 GB)

Web upload has no hard size cap, but a browser upload gets slow above ~2 GB and unreliable above ~5 GB, and it cannot resume: if the tab closes mid-upload, the transfer aborts. Above ~2 GB, switch to the SuperRenders Client App, or use S3 access (Cloud Direct Connect) with Cyberduck or the AWS CLI as described in Upload and download. The Client App is resumable, so a dropped connection doesn't mean starting the transfer over, and it uploads in parallel chunks, which cuts total time on large asset libraries.

How long your files stick around

Rendered output and project files stay available for download after the job finishes; there is no fixed automatic deletion period, and we delete a project when you ask us to. Finished renders come back through a direct web download, automatic download through the Client App if you've set that up, or an S3 pull from SuperRendersOutput/<job ID>/ with Cyberduck or the AWS CLI.

In practice: download your output promptly, or set up Client App auto-download so it lands on your local drive without you having to remember. Keep your own backup of the original scene files; the farm is not an archive.

Cross-references

Missing textures or assets in your farm render: causes and fixes — what to check before you upload if your scene depends on external assets
Supported versions and plugins matrix — confirm your DCC, render engine, and plugin versions are covered
Contact support — if you're not sure a file or format will work, we can confirm before you upload

FAQ

Q: Can I upload a .zip? A: You can upload one, but the farm does not extract archives (.zip, .rar, .7z, .tar, .tar.gz), so nothing inside it gets rendered. Extract it on your workstation and upload the folder unpacked.

Q: Can you deliver renders to my Google Drive or Dropbox? A: No. Finished renders come back through web download, Client App auto-download, or an S3 pull with Cyberduck or the AWS CLI; move them to your cloud storage yourself if that's part of your workflow.

Q: Is there a file size limit if I use the Client App? A: No practical limit. We recommend it for any project above ~2 GB because it supports resumable, parallel transfers, while a browser upload gets slow above ~2 GB, becomes unreliable above ~5 GB and cannot resume.

Updated on: 29/09/2026

Was this article helpful?

Thank you!