What visitors experience

Your lens page is the visitor view

Open Lens in the sidebar and you are looking at exactly what your visitors see: the same greeting, the same answers, replied through the real pipeline with citations included. If it refuses something it should not, you see that too, before your audience does.

Visitors on your public page see the same content presented as a centered card; only the owner view goes edge to edge. Use the open-in-new-tab arrow in the top bar to view the public page exactly as a signed-out visitor.

Test chats stay private

Conversations you have with your own lens never show up in your Inbox, Audience, or Analytics. They do not fire webhooks or notifications either. Test as much as you like; your stats stay honest.

Your test conversation is saved, so you can pick it up where you left off on any device. Use Start over in the chat menu to clear it.

Visitors can attach images and files

When someone chats with your lens, they can attach images (JPG, PNG, WebP, GIF) and documents (PDF, TXT, Markdown, CSV, Word) alongside their question. The lens actually sees the image or reads the document when it answers; it is not just decoration.

1. The visitor taps the paperclip in the message box and picks one or more files, or pastes or drops an image.
2. Each file shows a small preview in the box with a loading spinner while it is prepared. Nothing is sent yet.
3. The visitor can add or remove files, type a message, then press Send. The text and attachments go together in one turn.

Images are compressed in the visitor’s browser before sending, so uploads stay fast. Up to 4 files can be attached at once.

Good to know:

• Attachments run on your plan, the same as any lens question, and count toward your daily question limit.
• If your plan does not include image or document uploads, those attachments are quietly skipped and the lens still answers the text.
• Very large files are rejected in the message box before sending, with a short note to the visitor.

Teach it by dropping files

As the owner, drag a PDF, audio, video, image, or markdown file anywhere onto the chat. A training bubble appears in the conversation while your lens reads it:

• Reading - the file is uploading.
• Learning - your lens is absorbing it.
• Learned N things - ask me about it - done; ask a question about the file right away.

Everything you drop lands in this workspace’s knowledge, visible to this lens.

For everything else, use the Teach button in the top bar. It opens the same composer as Knowledge: write or paste anything (links and YouTube videos are detected automatically), attach files, or tap the mic and just say it, and your words become editable text before you save. When you press Enter, the panel closes and a training bubble in the chat narrates the learning.

The top bar

• Live / Paused - pause your lens so visitors cannot reach it. You can keep chatting with it here while it is paused, and resume it anytime.
• The link chip - click to copy your public link.
• Share - QR code, embed snippet, social card, custom link.
• Customize - every lens setting on one calm page; see Lens settings.
• The open arrow - open the public page in a new tab to see it exactly as a signed-out visitor.

Try this

1. Open Lens and ask it a question a customer would ask.
2. If the answer is thin, drop the document that has the answer into the chat.
3. Wait for “Learned - ask me about it,” then ask again.
4. Then try the visitor side: tap the paperclip, add a photo, and ask “What’s in this picture?”