Overview
Media commands give your AI agent direct access to files inside any chat. Type /, drop in an image, PDF, or audio clip, and the agent reads it, transcribes it, or extracts the data — no separate upload step, no leaving the conversation.
TL;DR: Slash commands turn agent chat multimodal. Read images, transcribe audio, extract from PDFs, analyze video frames — all in the same thread, all aware of the agent's existing knowledge base.
Using Media Commands
- Open a conversation with any AI agent.
- Type
/to see the list of available slash commands. - Select a media command to upload or reference a file.
- Your agent processes the file and responds with relevant analysis.
Available Commands
| Command | What It Does |
|---|---|
| Upload media | Attach images, PDFs, or documents to the conversation |
| Reference media | Point your agent to existing files in your workspace |
| Analyze media | Ask your agent to extract information from uploaded files |
Use Cases
- Analyze screenshots — Upload a screenshot and ask your agent to extract text or identify issues
- Process documents — Upload a PDF and ask for a summary, key points, or action items
- Work with images — Upload product photos and ask your agent to write descriptions or captions
- Review designs — Share design mockups and get feedback from your agent
Combining with Agent Knowledge
Media commands work alongside your agent's knowledge base. When you upload a file in chat, the agent can cross-reference it with its existing knowledge for richer, more contextual responses.
For example: Upload a client proposal → your agent compares it against your pricing database and suggests adjustments.
Mini FAQ
Which file types are supported? Images (PNG, JPG, GIF, WebP), PDFs, audio, video, and most common document formats. The agent picks the right tool automatically.
Does the file get added to long-term knowledge? No — files in chat are scoped to that conversation. To make them permanent, add them to the agent's knowledge base.
Do media commands burn extra credits? Vision and transcription calls use credits per file. See AI Usage for details.
Related guides
- AI Agent Tools — Full list of built-in agent tools
- Agent Knowledge — Train agents with custom data
- Taskade EVE Mentions — Reference workspace items in conversations
- Media Tab — Manage media files in your workspace
- Media Chat — Chat directly with uploaded files