Chat Tools Menu

The + menu in the chat composer adds extra context to a message. Available tools depend on platform, permissions, provider support, and the selected model.

Common tools

• Camera: capture a new image for a vision-capable model.
• Photos: choose images from your photo library.
• Web / Jina Reader: fetch readable content from a URL.
• File: add text, code, Markdown, CSV, JSON, and similar files.
• PDF: extract PDF text or convert pages for vision workflows.
• Location: share a map location with the model.
• MCP tools: enable configured tools for supported LM Studio v1 chats.

What gets sent

Tools add content to your message. That content is sent to the selected model or provider when you send the message. Token estimates help you understand how much context the tool content may use.

Tips

• Use one or two focused attachments instead of many large files.
• Choose a vision-capable model before sending images.
• Use Local Models for private on-device text and supported local vision workflows.
• Use remote providers when you need larger context, faster image analysis, or server-side tools.

Troubleshooting

• If a tool is disabled, check the selected model and platform permissions.
• If sending fails, remove large attachments or lower token usage.
• If web content fails, confirm the URL is public and reachable.
• If location is unavailable, check iOS location permission for 3sparks Chat.