Overview
You can pass images to ChatGPT in Zapier using the Analyze Images action or the Conversation action, both powered by OpenAI's Responses API. These actions let you upload images, then send a message or question about them. This is useful for workflows that need to describe, classify, extract data from, or moderate image content automatically.
Step-by-step guide
Method 1: Use the Analyze Images action
- Add a ChatGPT (OpenAI) action to your Zap. In the Zap editor, click the + icon to add an action step. Search for and select ChatGPT (OpenAI).
- Select Analyze Images as the action event. Click the Action event dropdown and choose Analyze Images. Click Continue.
- Connect your OpenAI account. Select an existing connection or add a new one using your OpenAI API key. Click Continue.
- Add your image in the Images field. In the Images field, map an image file or URL from a previous step (such as an email attachment or Google Drive file trigger) in your Zap. The image must be in a supported format (JPEG, PNG, GIF, or WebP) and must be publicly accessible if you are passing a URL.
- Write your prompt in the Message field. Tell ChatGPT what you want it to do with the image. For example: "Describe what is in this image", "Extract any text visible in this image", or "Classify this product image into one of these categories: electronics, clothing, furniture".
- Select a model. Choose a vision-capable model such as gpt-4o. Only models that support image input will process the Images field.
- Test and use the output. Click Test step. ChatGPT will return a text response about the image, which you can map to subsequent steps in your Zap.
Method 2: Use the Conversation action
- Add a ChatGPT (OpenAI) action and select Conversation. This is the recommended general-purpose action powered by the Responses API. It supports image input alongside text.
- Add your image in the Images field. In the Images field, map an image file or URL from a previous step (such as an email attachment or Google Drive file trigger) in your Zap. The image must be in a supported format (JPEG, PNG, GIF, or WebP) and must be publicly accessible if you are passing a URL.
- Select a vision-capable model. Choose any model that supports image input. Check the model selector for vision support.
- Test the step. ChatGPT will respond with text based on both your message and the image content.
Key notes and limitations
- Images passed via URL must be publicly accessible. If your image is stored in a private location (such as a private Google Drive file or a Dropbox shared link), you will need to generate a direct, public download URL before passing it to ChatGPT.
- Supported image formats are JPEG, PNG, GIF, and WebP.
- The Images field will also accept data URIs (e.g., data:image/jpeg;base64,...).
- Images must be under 20 MB in size.
- You need an OpenAI account with prepaid API billing enabled. A ChatGPT subscription (Plus, Pro) does not give you API access, these are separate.
Frequently asked questions
Q: How do I upload a photo to ChatGPT with a Zap?
A: In the Images field of the ChatGPT action, map the photo file or URL from any previous step that outputs an image or image URL.
Q: Can I upload multiple images to ChatGPT in one Zap step?
A: Yes. The Analyze Images and Conversation actions support multiple images. Add each image as a separate entry in the Images field by clicking the + button to add additional image inputs.
Q: Can I send multiple images to ChatGPT in a ZIP file?
A: Yes. You can map a ZIP file containing multiple images to the Images field and they will be automatically extracted and passed to the ChatGPT.
Q: Why is my ChatGPT image upload not working in Zapier?
A: The most common causes are either the image URL is not publicly accessible, the file format is not supported (only JPEG, PNG, GIF, WebP), the file is over 20 MB, or a text-only model has been selected. Check each of these before testing again.
Q: Can ChatGPT generate an image via Zapier?
A: Not through the Analyze Images or Conversation actions, these return text only. To generate an image from a text prompt, use the Generate an Image action.
Q: Can I use Zapier to edit an existing image with ChatGPT?
A: Yes. Use the Create Image Edit action in the ChatGPT (OpenAI) app. This action takes an existing image and a text prompt describing the changes you want, and returns a new edited version of the image. It is separate from the Generate An Image action, which generates images from a text prompt only.