Importing videos

TubePress can fill your library in several ways — from adding a single clip by hand to pulling in hundreds of thousands of ready-made videos from the cloud. This page covers the built-in importer and shows you where each other route lives.

Whichever method you choose, the result is the same: ordinary videos you can edit, organise, moderate and publish like any other. Pick the one that matches how much content you have, and whether you want a one-off load or something that keeps topping itself up.

Ways to get content into TubePress

There is no single "right" way to add videos — each method suits a different scale and workflow. Use the table below to choose, then follow the link for the details.

Importing a single video

TubePress has no separate "one video" importer — to add a single video you use the Videos screen at /admin/videos and choose Add Video. There are two ways to provide the media:

• Manual upload — upload a video file you host yourself.
• Single-URL import — paste an <iframe> embed code or a direct video URL to embed a clip hosted elsewhere.

Fill in the title and thumbnail, attach any categories, tags or performers, then save as a draft or publish straight away. Neither route uses credits.

Bulk import

Bulk import lives on the Upload File tab of /admin/import. It takes a single CSV or JSON file — not a pasted list of URLs — and turns each row or object into a video.

• CSV must have a header row and use commas as separators.
• JSON must be an array of objects.

Drag the file onto the upload zone (or click to browse) and TubePress reads the columns and shows a preview. A minimal CSV looks like this:

title,embed_code,thumbnail,duration,tags
My clip,https://ex.com/v/123.mp4,https://ex.com/t.jpg,612,"solo,hd"

Next you map your columns onto TubePress fields by dragging them into place. The fields the importer recognises include:

Before you start, set how duplicates are handled (skip or update, matched on video_id), whether to download thumbnails so they are served from your storage, and any optional AI rewriting or translation.

How the import queue works

For anything beyond a small file, choose Start in background. TubePress creates a job and hands it to the import processor, which works through your rows even after you close the tab. No system crontab is required — it runs on the built-in pseudo-cron.

The processor is designed to be safe and resilient:

• Chunked. Rows are imported a few at a time so a web worker is never tied up; an adjustable import speed controls how many per chunk.
• Asynchronous. The pseudo-cron picks the job up as the site receives traffic, and the import page nudges it along while open.
• Self-healing. A job that stalls — for example a server restart mid-run — is requeued automatically after about five minutes and continues where it left off.
• De-duplicating. Rows with a known video_id are skipped or updated rather than duplicated.
• Forgiving. A bad row is logged (the first few errors per chunk are sampled) and the job carries on.

Watching import progress

Open /admin/import and watch the Import Queue. Each active job shows live counts — imported, skipped, errors, and processed-of-total — refreshed every few seconds.

From here you can pause, resume, cancel or delete a job. Jobs from the last 24 hours stay listed so you can review what happened, including sampled error messages for any rows that failed.

Tips & limits

• Only .csv and .json files are accepted; other formats are rejected up front.
• The practical minimum per row is a title plus an embed code and a thumbnail.
• Upload size is capped by your server's upload_max_filesize and post_max_size. Split very large files, or rely on a background job.
• Map a stable video_id so re-running a file updates rather than duplicates.
• Enable Download thumbnails to self-host images instead of hotlinking the source.
• AI rewriting and translation consume credits and need a linked tubepress.io account; if your balance is too low the job stops with a clear message.

Next steps