Skip to content

打标助手

Batch caption images for model training data

Visit official site

What 打标助手 is

The tool is described as an online tool built on GPT-4 Vision that uses tuned prompts to add text captions to batches of images. The captions are presented as labelled data for training SD models.

What you can do with it

  • Download caption files in batches as .txt files named after images
  • Add or remove a caption prefix across a batch
  • Edit captions by hand to correct automatic results
  • Send uploaded images straight to OpenAI for processing

Application scenarios

  • Researchers prepare labelled data for model training
  • Data scientists process experiment images in batches
  • Business users caption a product catalogue for search optimisation

Getting started

  • Visit the site and enter a GPT-4 Vision API key
  • Make sure the key works in the OpenAI account
  • Upload the images that need captions
  • Set an image prefix for later management

Pros & cons

✓ What we like

  • Captions can be edited by hand to fix errors
  • Uploaded images go straight to OpenAI for processing

! What to watch out for

  • A GPT-4 Vision API key is required
  • Use is limited to 100 calls per day

FAQ

What is needed before uploading?

The steps name entering a GPT-4 Vision API key that works in an OpenAI account.

What happens to uploaded images?

The material says images are sent directly to OpenAI and that the tool does not access or store uploaded data or the API key.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review