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

# AI Chat

> Chat with AI models to get help with any task

## Overview

EnConvo provides powerful AI chat capabilities that let you converse with various AI models for assistance with any task. Whether you need help writing, coding, researching, or problem-solving, AI Chat has you covered.

## Supported AI Models

EnConvo supports a wide range of AI providers and models:

### Cloud Models

| Provider      | Models                                           |
| ------------- | ------------------------------------------------ |
| **OpenAI**    | GPT-4o, GPT-4, GPT-3.5 Turbo                     |
| **Anthropic** | Claude 3.5 Sonnet, Claude 3 Opus, Claude 3 Haiku |
| **Google**    | Gemini Pro, Gemini Ultra                         |
| **Mistral**   | Mistral Large, Mistral Medium                    |

### Local Models (Privacy-First)

| Platform      | Models                                |
| ------------- | ------------------------------------- |
| **Ollama**    | Llama 3, Mistral, CodeLlama, and more |
| **LM Studio** | Any GGUF model                        |

## Chat Features

### Multi-Turn Conversations

Maintain context across multiple messages for natural, flowing conversations.

### Context Integration

Automatically include:

* Screen content
* Selected text
* Clipboard data
* Knowledge base references
* File attachments

### Response Handling

* **Copy**: Copy response to clipboard
* **Regenerate**: Get a different response
* **Edit & Resend**: Modify your prompt
* **Continue**: Ask for more details
* **TTS**: Listen to response (Text-to-Speech)

### Voice Input In Chat Window

Use the microphone button in the Chat Window prompt to dictate a message instead
of typing. This uses the same dictation provider configured under
**Settings → Dictation**, so you can choose a real-time cloud provider, a
subscription provider, or an offline local model.

<Tip>
  Voice input is best for natural follow-up prompts such as "summarize this in
  three bullets" or "turn that into a shorter email."
</Tip>

### Conversation Navigation

The Chat Window keeps conversation history in sync with the sidebar and titlebar
controls:

* Use the titlebar back and forward buttons to move through recent chat views.
* Use the history sidebar to switch conversations.
* Use row actions to manage individual history items when available.
* Use prompt mentions to add tools, apps, websites, files, or Knowledge Base
  context before sending.

### Update Button

When EnConvo detects a newer version for your selected update channel, the Chat
Window can show an **Update** button in the titlebar. Click it to open the
update flow.

For update channels and manual checks, see [Updates](/configuration/updates).

## Chat with Documents

Upload documents directly into chat for AI analysis:

<Steps>
  <Step title="Attach Document">
    Drag & drop a file into chat, or click the attachment icon
  </Step>

  <Step title="Supported Formats">
    PDF, Word (.docx), Text (.txt), Markdown (.md), Code files
  </Step>

  <Step title="Ask Questions">
    "Summarize this document"
    "What are the key points?"
    "Find information about X"
  </Step>
</Steps>

## Chat with Webpages

Analyze and discuss webpage content:

<Steps>
  <Step title="Include Webpage">
    Use `@webpage` or paste a URL in your chat
  </Step>

  <Step title="Ask Questions">
    "Summarize this article"
    "What is this page about?"
    "Extract the main arguments"
  </Step>
</Steps>

## Text-to-Speech (TTS)

Have AI responses read aloud:

### Enabling TTS

1. Click the speaker icon on any AI response
2. Or enable auto-TTS in Settings → Chat → Text-to-Speech

### Voice Options

* Multiple voice options
* Adjustable speed (0.5x - 2x)
* Different languages
* Male/Female voices

## Model Selection

### Switching Models

Click the model name in the chat header to switch between configured models.

The model picker keeps spacing compact and applies your selection to the current
conversation. If a model requires a provider login or API key, configure that
provider first under **Settings → AI Provider**.

### Model-Specific Features

| Model                 | Best For                    |
| --------------------- | --------------------------- |
| **GPT-4o**            | Complex reasoning, analysis |
| **Claude 3.5 Sonnet** | Writing, nuanced tasks      |
| **Gemini Pro**        | Multimodal, long context    |
| **Local LLMs**        | Privacy, offline use        |

### Custom Model Settings

Configure per-model settings:

* Temperature (creativity level)
* Max tokens (response length)
* System prompt
* Top-p sampling

## Chat Commands

Use these commands in chat for quick actions:

| Command         | Action                     |
| --------------- | -------------------------- |
| `/new`          | Start new conversation     |
| `/clear`        | Clear current conversation |
| `/model [name]` | Switch AI model            |
| `/context`      | Show current context       |
| `/export`       | Export conversation        |
| `/help`         | Show available commands    |

## Keyboard Shortcuts

| Shortcut  | Action             |
| --------- | ------------------ |
| `⌘ Enter` | Send message       |
| `⇧ Enter` | New line           |
| `⌘ N`     | New conversation   |
| `⌘ L`     | Clear conversation |
| `↑`       | Edit last message  |
| `⌘ C`     | Copy last response |

## Message Rendering

Chat messages support Markdown, code blocks, tables, images, files, and tool
results. Long tables are scrollable so they do not break the conversation
layout, and text selection is optimized for copying only the content you need.

## Configuration

### Settings → Chat

| Setting              | Description                 |
| -------------------- | --------------------------- |
| **Default Model**    | Model used for new chats    |
| **Temperature**      | Response creativity (0-1)   |
| **Stream Responses** | Show text as it generates   |
| **Auto-TTS**         | Read responses aloud        |
| **History Length**   | Messages to keep in context |
| **System Prompt**    | Custom instructions for AI  |

## Tips for Better Responses

<AccordionGroup>
  <Accordion title="Be specific">
    Instead of "Write about AI", try "Write a 500-word blog post about how AI is transforming healthcare in 2024, focusing on diagnosis and treatment planning."
  </Accordion>

  <Accordion title="Provide context">
    Share relevant background: "I'm a software developer working on a React app. How do I..."
  </Accordion>

  <Accordion title="Use examples">
    Show the AI what you want: "Format the response like this: \[example]"
  </Accordion>

  <Accordion title="Break complex tasks">
    Divide large requests: "First, outline the structure. Then, write section 1..."
  </Accordion>
</AccordionGroup>

## Related Features

<CardGroup cols={2}>
  <Card title="Knowledge Base" icon="database" href="/ai/knowledge-base">
    Chat with your documents
  </Card>

  <Card title="Local LLM" icon="shield-halved" href="/ai/local-llm">
    Privacy-first local models
  </Card>
</CardGroup>