Translating your Figma frames with ChatGPT

Use ChatGPT inside CopyDoc with either direct API translation or manual CSV prompts to localize your selected Figma frames in Figma.

CopyDoc supports ChatGPT in two ways:

  • ChatGPT API for automated translation and import inside the plugin.
  • ChatGPT Prompt for a manual copy and paste workflow using the ChatGPT app.

Use ChatGPT API ​

The API workflow is the fastest option if you want CopyDoc to handle the full translation and import flow for you.

  1. Click the Localize Text button in CopyDoc.
  2. Choose ChatGPT API.
  3. Paste your OpenAI API key.
  4. Pick the model you want CopyDoc to use.
  5. Select the Figma frames and target locales you want to translate.
  6. Enable Use Markdown if you need to preserve inline bold, italic, and linked text during translation.
  7. Click Translate & Import.

CopyDoc will send the selected frame text to ChatGPT for each locale, then automatically import the translated results back into Figma as cloned localized frames.

Your ChatGPT API key is stored per Figma user on your device and is not saved into the Figma file itself.

Use ChatGPT Prompt ​

The manual prompt workflow is useful if you prefer to run translation requests yourself inside ChatGPT instead of connecting an API key.

  1. Choose ChatGPT Prompt in the localization options.
  2. Select the Figma frames and locales you want to translate.
  3. Click Generate Prompt.
  4. Click Copy Prompt to Clipboard.
  5. Run the prompt in ChatGPT and make sure it returns raw CSV only.
  6. Paste the CSV response back into CopyDoc.
  7. Click Translate Figma Frames to import the translated content into Figma.

In the manual prompt workflow, you choose the ChatGPT model inside the ChatGPT app yourself. CopyDoc validates the pasted CSV before importing it.

The generated prompt also tells ChatGPT to preserve supported inline markdown formatting and keep markdown markers, URLs, and placeholders unchanged so the translated content can be imported back into Figma correctly.

Do not remove or rename the id, frame_id, layer_name, or figma_text columns in the CSV response. CopyDoc uses those columns to match translated rows back to the correct Figma layers.