> ## Documentation Index
> Fetch the complete documentation index at: https://docs.relayhub.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat Basics

> Learn how to chat with AI models in RelayHub

## The Chat Interface

RelayHub's chat interface is designed to feel familiar while giving you powerful AI capabilities. The left sidebar shows your conversation history, and the main area is where you interact with the AI.

<Columns cols={3}>
  <Card title="Sidebar" icon="bars">
    Browse, search, and manage your conversations. Pin important ones to the top.
  </Card>

  <Card title="Message Area" icon="message">
    View the full conversation thread with streaming AI responses.
  </Card>

  <Card title="Settings Bar" icon="sliders">
    Select your model, toggle features, and attach files -- all without leaving the chat.
  </Card>
</Columns>

## Starting a Conversation

<Steps>
  <Step title="Open a new chat">
    Click the **New Chat** button in the sidebar or use the keyboard shortcut. A fresh conversation opens with your default AI model selected.
  </Step>

  <Step title="Choose your AI model">
    Use the model selector at the top of the chat to pick which AI model you want to talk to. RelayHub supports models from OpenAI, Anthropic, Google, and xAI. Each model has different strengths -- hover over a model name to see a brief description.
  </Step>

  <Step title="Type your message">
    Enter your message in the input area at the bottom of the chat. Press **Enter** to send, or **Shift+Enter** to add a new line.
  </Step>

  <Step title="Watch the response stream in">
    AI responses stream in token by token, so you see results immediately rather than waiting for the full answer. You can stop generation at any time by clicking the **Stop** button.
  </Step>
</Steps>

## Selecting AI Models

RelayHub gives you access to multiple AI providers and models. You can change the model at any point in a conversation, and your choice persists for that conversation.

<Tabs>
  <Tab title="OpenAI">
    OpenAI models excel at general-purpose tasks, coding, and following detailed instructions. Available models range from fast and affordable (GPT-5-nano) to highly capable (GPT-5.3).
  </Tab>

  <Tab title="Anthropic">
    Anthropic's Claude models are known for careful reasoning, long document analysis, and nuanced writing. Choose Claude Sonnet for everyday tasks or Claude Opus for complex reasoning.
  </Tab>

  <Tab title="Google">
    Google's Gemini models offer strong multimodal capabilities and fast response times. Gemini Flash models are optimized for speed, while Pro models handle complex tasks.
  </Tab>

  <Tab title="xAI">
    xAI's Grok models provide strong reasoning and vision capabilities. Grok 4 handles complex multi-step tasks, while Grok 3 is a solid general-purpose choice.
  </Tab>
</Tabs>

<Tip>
  Not sure which model to pick? Start with a mid-tier model for everyday questions. Switch to a more powerful model when you need deep reasoning or complex analysis.
</Tip>

## Conversation History

Every conversation is automatically saved. You can find past conversations in the sidebar, sorted by most recent. RelayHub auto-generates a title for each conversation based on the first few messages, so you can scan your history at a glance.

To manage conversations:

* **Rename** -- Click the three-dot menu on any conversation and select **Rename** to give it a custom title.
* **Delete** -- Click the three-dot menu and select **Delete**. This is permanent and cannot be undone.
* **Search** -- Use the search bar at the top of the sidebar to find conversations by title or content.

<Warning>
  Deleting a conversation removes it for all participants if the conversation was shared. Make sure no one else needs it before deleting.
</Warning>

## Chat Settings

Each conversation remembers its own settings independently. These persist across sessions, so when you return to a conversation days later, your configuration is exactly as you left it.

| Setting            | What It Does                                               |
| ------------------ | ---------------------------------------------------------- |
| **Model**          | Which AI model responds to your messages                   |
| **Web Search**     | Lets the AI search the internet for current information    |
| **Reasoning Mode** | Shows the AI's chain-of-thought process before answering   |
| **RAG Mode**       | Searches your uploaded knowledge base for relevant context |

<Note>
  Settings changes take effect on the next message you send. They do not retroactively change previous responses in the conversation.
</Note>

## Streaming Responses

RelayHub streams AI responses in real time. As the model generates text, you see it appear word by word. This means you can start reading while the AI is still thinking, and you can stop generation early if the response is heading in the wrong direction.

If the AI's response is not what you expected, click the **Regenerate** button to get a new response from the same model using the same message.