How to Bulk-Generate Alt Text From a CSV in 5 Minutes
If you already have a spreadsheet of product or content images, you don’t need to upload them one by one. This walkthrough covers the CSV import path from start to finish.
1. Prepare your CSV
Your file needs one required column:
| Column | Required | Description |
|---|---|---|
url | Yes | Full URL to the image |
page_title | No | Title of the page the image is on |
product_name | No | Name of the product |
brand_name | No | Brand or manufacturer |
page_description | No | Meta description |
nearby_text | No | Surrounding text for extra context |
The optional columns aren’t required, but they help the model write a more accurate description when you have them. You can grab a ready-made template from the Import & Export page inside AltGen.
2. Upload the file
Go to Import & Export → Import, drop your CSV, and AltGen validates the file and shows you a preview before anything is queued. You’ll catch a malformed URL or missing column here instead of after the batch runs.
3. Configure the batch
Pick:
- Style — Terse, Concise, Direct, or Elaborated
- Preset — Ecommerce, Real Estate, Travel, and others tuned for how the description should read
- Languages — Generate in up to 30 languages in the same batch, side by side
4. Run it
AltGen queues every row and processes the batch in the background. A single batch can hold up to 5,000 images, and failed rows retry automatically instead of silently dropping out.
5. Review and export
Once the batch completes, everything lands in the Image Library with a status per image. Edit anything that needs a human touch, then export back to CSV or pull results through the API.
That’s the whole workflow: spreadsheet in, accessible and searchable alt text out, without opening a single image individually.