Chat Interface Guide

The Vectly chat interface is designed for productive AI conversations with powerful features and intuitive controls. Learn how to make the most of every interaction.

Interface Overview

Main Components

1. Model Selector (Top Bar)

• Current Model Display: Shows active AI model
• Quick Switch: Click to change models mid-conversation
• Credit Indicator: Real-time credit usage display

2. Chat Area (Center)

• Message History: Scrollable conversation view
• Streaming Responses: Real-time AI responses
• Message Actions: Copy, regenerate, or edit messages

3. Input Area (Bottom)

• Message Composer: Multi-line text input
• File Attachment: Upload documents for context
• Send Button: Submit your message
• Feature Toggles: Quick access to web search, reasoning

Starting a Conversation

Creating a New Chat

1. Click "New Chat" in the sidebar or press Ctrl/Cmd + N
2. Select your preferred AI model
3. Type your message and press Enter

Chat Settings

Access settings via the gear icon:

• Model Selection: Choose or change AI model
• Advanced Features: Enable web search, reasoning, caching
• Context Files: Manage uploaded documents
• Chat Title: Rename for easy identification

Message Features

Sending Messages

Basic Messages

• Type your message in the input box
• Press Enter to send (Shift+Enter for new line)
• Messages appear instantly with streaming responses

Multi-line Messages

• Use Shift+Enter to create line breaks
• Format with markdown for better readability
• Include code blocks with ``` syntax

Message Formatting

Vectly supports full markdown:

• Bold: **text**
• Italic: *text*
• Code: `code`
• Lists, quotes, headers, and more

Message Actions

For Your Messages

• Edit: Click to modify sent messages
• Delete: Remove from conversation
• Copy: Copy message text

For AI Responses

• Copy: Copy the full response
• Regenerate: Get a new response
• Continue: Have AI continue from where it stopped
• Report: Flag problematic responses

Advanced Features

Web Search

Enable AI to search the internet:

1. Toggle the search icon in the input area
2. AI will search when relevant to your query
3. See search results inline with citations
4. Additional credits apply for searches

Reasoning Mode

Available for supported models:

• Visible Reasoning (Claude): See AI's thought process
• Reasoning Effort (O1): Adjust computation depth
• Useful for complex problems and analysis

Prompt Caching

Reduce costs for repetitive contexts:

• Automatically enabled for Claude models
• Caches common prompt prefixes
• Significant savings for long conversations
• See cache hits in credit usage

File Attachments

Add context to your conversations:

1. Click the paperclip icon
2. Select files (PDF, TXT, MD supported)
3. Files are processed and indexed
4. AI references content automatically

Conversation Management

Chat Organization

Automatic Saving

• All chats save automatically
• Resume anytime from the sidebar
• Search through chat history
• No manual save needed

Chat Titles

• Auto-generated from first message
• Click title to rename
• Use descriptive names for easy finding

Moving Chats

• Assign chats to projects
• Drag and drop in sidebar
• Bulk operations available

Context Management

Conversation Memory

• AI maintains context within a chat
• Previous messages influence responses
• Clear context with "New Chat"

Context Window

• Each model has token limits
• Older messages may be truncated
• Important context should be repeated

Productivity Tips

Keyboard Shortcuts

Best Practices

1. Clear Communication

• Be specific in your requests
• Provide context upfront
• Use examples when helpful
• Break complex tasks into steps

2. Effective Prompting

• Start with the end goal
• Specify format preferences
• Include relevant constraints
• Ask for clarification if needed

3. Context Usage

• Upload relevant documents
• Reference previous messages
• Use project-level context
• Clear context when switching topics

4. Credit Optimization

• Start with cheaper models
• Enable caching for long contexts
• Be concise but clear
• Monitor usage in real-time

Special Features

Code Handling

• Syntax highlighting for 100+ languages
• Copy button for code blocks
• Line numbers for reference
• Automatic language detection

Table Support

• Markdown table rendering
• Copy as formatted text
• Export to CSV (coming soon)
• Responsive design

Math Rendering

• LaTeX support for equations
• Inline and block math
• Copy as LaTeX or image
• Professional formatting

Troubleshooting

Common Issues

Slow Responses

• Check model selection (some are slower)
• Verify internet connection
• Try a faster model like GPT-4o Mini

Context Errors

• Model context limit reached
• Start a new chat
• Remove unnecessary context
• Use a model with larger context window

Formatting Problems

• Use proper markdown syntax
• Preview before sending
• Check for unclosed code blocks
• Report persistent issues

Tips & Tricks

Power User Features

1. Template Messages: Save common prompts
2. Message History: Use up/down arrows
3. Quick Copy: Double-click code blocks
4. Focus Mode: Hide sidebar with Ctrl+/

Coming Soon

• Voice input and output
• Image generation
• Multi-chat view
• Collaborative editing
• Custom themes

Need more help? Contact [email protected]