LabelAssay Alcohol Beverage Label Check · 27 CFR
← Back

Check Many Labels

Up to 1000 files or 1600 MB per upload, each image 5 MB or less. Add a CSV of application data to check each label against its own row. Results appear as they finish.

Label images are sent to the configured AI reader (currently Anthropic's Claude API) to read the printed text, with camera metadata (EXIF, including location) removed first; they are not stored after the check.

Columns: filename, brand_name, class_type, and optionally fanciful_name; when a fanciful name is filed, the brand check accepts either name. Without the CSV, the brand check is skipped; all other checks still run.

Re-reads the label rotated, up to three extra passes for that label; uncheck for maximum throughput.

Uploading the batch — large drops take a moment.