Skip to content

Attachments

You can send images and audio files alongside your messages. When the AI supports multimodal input (vision or audio), it can see and respond to your attachments.

Sending an Attachment

  1. Click the Attachment button in the input area (or drag a file into the chat)
  2. Select an image or audio file
  3. The attachment appears as a preview in the input area
  4. Type your message (optional) and send

The attachment is uploaded to Lumiverse's image system and stored alongside the message.

Supported Formats

Type Formats
Images PNG, JPG, WebP, GIF
Audio WAV, MP3, and other common audio formats

How the AI Sees Attachments

When you send a message with an attachment, Lumiverse converts the file to base64 and includes it as multipart content in the prompt. The AI receives both the text and the media together.

Model support

Not all AI models support vision or audio input. If your current model doesn't support it, the attachment is still saved with the message but may not influence the AI's response. Check your provider's model capabilities.

Provider-specific formatting is handled automatically:

  • OpenAI — Uses image_url and input_audio content parts
  • Anthropic — Uses image source blocks
  • Google — Uses inlineData parts

Viewing Attachments

Attachments appear inline in the message. Click on an image attachment to open it in the Image Lightbox for a full-size view.