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

# Chat

> See what you can do with your Chat

# Video Tutorial (coming soon)

<iframe style={{ borderRadius: '0.5rem', width: '100%', height: '400px'}} src="https://www.youtube.com/embed/6bF5p01MXqQ?si=kd45_8HryDQMIMcZ" />

# Text Tutorial

## Chat with your Expert

### Start to chat with your expert

Firstly click on the "Chat" Menu button on the left. Then click on the "Chat" button of your desired Expert.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/0ATpFtjrAtcClWI-/images/chat-first.png?fit=max&auto=format&n=0ATpFtjrAtcClWI-&q=85&s=0351068459e80aa183fb4c0d7d259eff" width="1920" height="945" data-path="images/chat-first.png" />

### The Chat window

After you clicked on the "Chat" button you will see the chat window. Here you can chat with your expert.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-second.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=8a242aeb4408d7b506971e3d541fb94b" width="1920" height="949" data-path="images/chat-second.png" />

## The Thumbbar

### The Thumbbar is the place where you can see all of the Chat functions

You can see the following functions in the Thumbbar:

* **Back Button**: Return to the Expert Overview
* **Abilities Button**: See and activate your expert's abilities
* **Chat History Button**: Browse previous conversations
* **Model Selection Button**: Change the AI model being used
* **Settings Button**: Configure chat settings (temperature, tokens, etc.)
* **Memories Button**: View and manage conversation memories
* **Tools & Connectors**: Enable/disable apps and integrations
* **Delete History**: Clear messages and start fresh
* **Save Chat**: Export conversation history
* **Upload Files**: Add documents, images, or other files
* **Voice Mode** 🎤: Enable voice input (new!)
* **DeepAgents Workspace**: Access todos and files (in Deep Mode)

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-third.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=c39b82478ecd65ec041fbd71b1e324c4" width="1920" height="948" data-path="images/chat-third.png" />

### The Abilities Window

You can see all of the abilities of your expert in the abilities window. You can see the name of the ability and the description of the ability.
Choose your desired ability with the "use Ability" to activate it.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-fourth.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=fcbf3b4dafa263140cda00fc1df978e1" width="1834" height="951" data-path="images/chat-fourth.png" />

### The Chat History

Search for a specific Chat in here and activate it by clicking on it.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-fifth.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=9185a178861ee11273e2f5fc7ade9f72" width="1920" height="951" data-path="images/chat-fifth.png" />

### The Model Selection Window

Choose your desired model in here, if you have trained an expert before you can choose between the models you have trained here.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-sixth.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=e6fc9f01bd539450b9455b05fbb2b6b3" width="1920" height="951" data-path="images/chat-sixth.png" />

### The Settings Window

Change your Settings in here:

* Temperature: The temperature of the model
* Max Tokens: The maximum tokens the model can use
* Top P: The top P value of the model
* Privat Session: If you want to have a private session with your expert, this will disable the tools of the expert and will use your private apps

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-seventh.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=1ed8a656d266a31a5258003be1379011" width="1920" height="843" data-path="images/chat-seventh.png" />

### The Tools and Connectors Window

You can see all of the tools and connectors of your expert in here. You can see the name of the tool and the description of the tool.
you can also disable or enable the tool in here.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-eight.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=cceb0ee38e89426cfc5d4161531e50e8" width="1920" height="846" data-path="images/chat-eight.png" />

## Chat Message functions

### The Chat Message functions are in the context popup of every message

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-ninth.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=30f7152b269ee8c6abe067ace27df7cd" width="1920" height="949" data-path="images/chat-ninth.png" />

### The Context Popup for a User message

by clicking on the three dots of a message you can see the following functions:

* Start QA Marking: Start the QA Marking for this Chat Messages
* Copy: Copy the message

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-tenth.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=1cbb806b631d1a08e7631821f52168bb" width="1920" height="946" data-path="images/chat-tenth.png" />

### The Context Popup for a Expert message

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-eleventh.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=26ee2a8783a4e5c375b65437b8281267" width="1920" height="951" data-path="images/chat-eleventh.png" />

by clicking on the three dots of a message you can see the following functions:

* Start QA Marking: Start the QA Marking for this Chat Messages
* Copy: Copy the message
* Thoughts: If the expert has thoughts about the message you can see them here, thoughts are sometimes tools or connectors the expert used for this message

<img style={{ borderRadius: '0.5rem' }} src="https://mintlify.s3.us-west-1.amazonaws.com/bbot-29a69ba9/images/chat-twelfth.png" />

### QA Marking

By enabling the QA Marking you can mark Messages for QA. These Marked QAs will be used to train the expert in the future.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-thirteen.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=91118c48de4f13890738bac981031186" width="1920" height="943" data-path="images/chat-thirteen.png" />

Here you can see the QA Marking Window. You can see the message you marked and the QA you assigned to it.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-sixteen.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=6a1d297686916e2cd6249a66903ff2ea" width="1920" height="949" data-path="images/chat-sixteen.png" />

### The Thoughts

If you see an Icon like this in the chat window you can see that the expert had thoughts on this message.
Thoughts are tools or connectors the expert used for this message.

<img style={{ borderRadius: '0.5rem' }} src="https://mintlify.s3.us-west-1.amazonaws.com/bbot-29a69ba9/images/chat-twelfth.png" />

Here you can see the thoughts of the expert. Thoughts are tools or connectors the expert used for this message.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-seventeen.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=d75e6f5240766a35134df5d8f92937bd" width="1859" height="933" data-path="images/chat-seventeen.png" />

There can be multiple thoughts for one message.

<img style={{ borderRadius: '0.5rem' }} src="https://mintcdn.com/bbot-29a69ba9/dm91uKYSfVL1m6HV/images/chat-eighteen.png?fit=max&auto=format&n=dm91uKYSfVL1m6HV&q=85&s=628ea00761b393bea2675a1d05948942" width="1920" height="842" data-path="images/chat-eighteen.png" />

## Voice Features

### Voice Input 🎤

Enable voice mode to speak to your expert instead of typing:

<Steps>
  <Step title="Enable Voice Mode">
    Click the **microphone icon** to toggle voice mode
  </Step>

  <Step title="Hold to Record">
    Press and hold the microphone button while speaking
  </Step>

  <Step title="Release to Send">
    Release when done - your speech is transcribed and sent automatically
  </Step>
</Steps>

**Voice Input Tips:**

* Speak clearly at a normal pace
* Minimize background noise
* Use a good quality microphone
* Voice is transcribed using AI (OpenAI Whisper or browser native)

### Audio Output 🔊

If text-to-speech is enabled, you'll hear your expert's responses:

* **Auto-play**: Responses play automatically
* **Manual play**: Click the play button to hear the response
* **Speed control**: Adjust playback speed
* **Voice selection**: Different voices available based on provider

**Supported TTS Providers:**

* OpenAI TTS (Alloy, Echo, Fable, Onyx, Nova, Shimmer)
* ElevenLabs (Premium voices)
* Google TTS (WaveNet voices)
* Azure Speech (Neural voices)
* Browser Native (Built-in)

Learn more: [Input & Output Modalities](/essentials/modalities)

## DeepAgents Workspace

### What is Deep Mode?

DeepAgents mode gives your expert a persistent workspace with tasks and files:

<CardGroup cols={2}>
  <Card title="Tasks Management" icon="tasks">
    Create, track, and complete todos
  </Card>

  <Card title="File System" icon="folder-open">
    Read, write, and manage files
  </Card>
</CardGroup>

### Enabling Deep Mode

<Steps>
  <Step title="Open Settings">
    Click the settings icon in the top bar
  </Step>

  <Step title="Enable Deep Mode">
    Toggle **"Deep Mode"** or **"Todo Mode"**
  </Step>

  <Step title="View Workspace">
    The workspace drawer appears on the right
  </Step>
</Steps>

### Using the Workspace

The workspace has two tabs:

**Tasks Tab:**

* View all todos created by the agent
* Track progress with status indicators
* Add manual tasks for the agent
* Complete or delete tasks
* Hierarchical task structure

**Files Tab:**

* View all files in the workspace
* Edit file contents
* Download files to your computer
* Upload new files
* Organize by type

**Task Statuses:**

* 🟢 **Completed**: Task finished successfully
* 🔵 **In Progress**: Agent is working on this
* ⚪ **Pending**: Not started yet
* 🔴 **Failed**: Encountered errors
* 🟠 **Cancelled**: Task was cancelled

Learn more: [DeepAgents Workspace](/essentials/deepagents-workspace)

## Multimodal Content

Your expert can handle multiple types of content:

<CardGroup cols={3}>
  <Card title="Text" icon="font">
    Traditional text messages
  </Card>

  <Card title="Voice" icon="microphone">
    Speak your questions
  </Card>

  <Card title="Images" icon="image">
    Upload and discuss images
  </Card>

  <Card title="Documents" icon="file">
    PDFs, Word docs, text files
  </Card>

  <Card title="Code" icon="code">
    Source code files
  </Card>

  <Card title="Data" icon="database">
    CSV, JSON, Excel files
  </Card>
</CardGroup>

### Combining Modalities

You can mix different input types in a single message:

**Example 1: Voice + Image**

```
[Voice] "What's in this image?"
[Image] screenshot.png
```

**Example 2: Text + File**

```
[Text] "Analyze this data and create a report"
[File] sales_data.csv
```

**Example 3: Multiple Files**

```
[Text] "Compare these documents"
[File] version1.pdf
[File] version2.pdf
```

## Advanced Chat Features

### HITL (Human-in-the-Loop) Mode

In Deep Mode with HITL enabled, the agent will ask for approval before taking actions:

**Auto Mode:**

```
Agent: ✓ Creating task...
Agent: ✓ Generating file...
Agent: ✓ Complete!
```

**HITL Mode:**

```
Agent: ❓ I propose to create 3 tasks and 2 files. Approve?
User: [Approve] [Reject] [Modify]
```

Toggle HITL in settings or say: *"Switch to HITL mode"*

### Private Session Mode

Enable private session to:

* Use your personal app credentials instead of expert's
* Keep conversations separate from expert's learning
* Override tool access

**When to use:**

* Accessing personal accounts
* Sensitive operations
* Testing without affecting expert
* Using different API keys

### Memory System

Your expert can remember information across conversations:

**Viewing Memories:**

1. Click the **brain icon** in the top bar
2. See all remembered facts and preferences
3. Edit or delete memories
4. Add new memories manually

**What gets remembered:**

* User preferences
* Important facts
* Project context
* Previous decisions

### Tool Events

When your expert uses tools, you'll see:

<CardGroup cols={2}>
  <Card title="Web Search" icon="magnifying-glass">
    Beautiful search result cards with titles, snippets, and relevance scores
  </Card>

  <Card title="API Calls" icon="plug">
    Clean display of API requests and responses
  </Card>

  <Card title="File Operations" icon="file-pen">
    Visual feedback for file reads/writes
  </Card>

  <Card title="Calculations" icon="calculator">
    Math and data processing results
  </Card>
</CardGroup>

**Expandable Tool Calls:**

* Click any tool call to expand/collapse
* See request details
* View full responses
* Copy data for your use

# Next Steps

<CardGroup cols={2}>
  <Card title="Train your Expert" icon="gauge" href="/essentials/training">
    Train your expert to make it smarter
  </Card>

  <Card title="Add Apps to your Expert" icon="messages" href="/essentials/apps">
    Add Apps to your expert to open more possibilities
  </Card>
</CardGroup>
