Who This Is For / When to Use
This feature is for teams using Conversation AI bots to handle customer inquiries that include photos, screenshots, or visual references.
Use this when:
Customers share product photos, properties, menus, or designs
Users send screenshots of errors, invoices, or confirmations
Conversations require visual context, not just text
What Image Response Does
When image responses are enabled, the AI bot can:
Analyze image content and context
Read visible text inside images using OCR
Combine image meaning with the customer’s message
Respond naturally using the bot’s existing prompt and knowledge base
No additional bot training is required.
Supported Image Types
Conversation AI supports common image formats:
JPG
JPEG
PNG
HEIC
Supported Channels
Image-based responses work automatically across:
Live Chat
WhatsApp
Facebook Messenger
Instagram
SMS (MMS)
How to Enable Image Responses
Step 1: Turn On Image Support in Bot Settings
Open your Conversation AI bot and go to Bot Settings.
Enable Also allow this bot to respond to: Images.
Once enabled, image handling is active for all supported channels.
Step 2: Test Image Responses
Send an image in a live conversation handled by the bot (for example, a photo sent via chat or messaging apps).
The bot will analyze the image and reply based on what it detects and understands.
How Image Understanding Works
Images are processed together with the user’s message
Visual context (objects, layout, scene) is analyzed
Text inside images is extracted using OCR
The bot uses its existing prompt, goals, and knowledge to respond
The response adapts automatically if the user corrects or replaces an image during the conversation.
Common Issues and Fixes
The bot does not respond to images
Confirm Images is enabled in Bot Settings
Ensure the conversation is handled by the AI bot (not a human-only flow)
Verify the channel supports MMS or image delivery
The bot misunderstands the image
Ask the user to clarify with a short message
Ensure the image is clear and not heavily blurred
Avoid extremely large or corrupted image files
Frequently Asked Questions
What happens if a user sends multiple images?
All images are analyzed together, and the bot responds based on the combined context.
Does the bot need separate training for images?
No. The bot uses the same prompt and knowledge base. Visual data is translated into context the bot already understands.
Can the bot read text inside images?
Yes. OCR allows the bot to read text from screenshots, receipts, invoices, menus, or handwritten notes when legible.
Are there image size or quality limits?
Standard mobile-friendly images work best. Extremely large, low-quality, or corrupted files may not process correctly.
Does this work with screenshots?
Yes. Screenshots of errors, forms, confirmations, or invoices are supported.
Which channels support image responses?
Live Chat, WhatsApp, Facebook Messenger, Instagram, and SMS (MMS).