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

# Anthropic Configuration

> Set up Anthropic Claude models in Glyph

Connect Glyph to Anthropic's API to use Claude 3.5 Sonnet, Claude 3 Opus, Claude 3 Haiku, and other models.

## Prerequisites

* Anthropic API account: [console.anthropic.com](https://console.anthropic.com)
* API key with appropriate permissions
* Sufficient API credits

## Setup

<Steps>
  <Step title="Get API Key">
    1. Log in to [Anthropic Console](https://console.anthropic.com)
    2. Navigate to **API Keys**
    3. Click **Create Key**
    4. Copy the key (starts with `sk-ant-`)

    <Warning>
      Store your API key securely. Anthropic only shows it once.
    </Warning>
  </Step>

  <Step title="Open Glyph AI Settings">
    Go to **Settings → AI** and select the **Anthropic** profile.
  </Step>

  <Step title="Add API Key">
    1. Click **Set API Key** in the authentication section
    2. Paste your Anthropic API key
    3. Click **Save**

    The key is stored in `.glyph/app/ai_secrets.json` in your space directory.
  </Step>

  <Step title="Select Model">
    Click the **Model** dropdown. Glyph fetches available models from Anthropic's API.

    Popular models:

    * `claude-3-5-sonnet-20241022` - Latest Claude 3.5 Sonnet (recommended)
    * `claude-3-opus-20240229` - Most capable, best for complex tasks
    * `claude-3-sonnet-20240229` - Balanced performance and speed
    * `claude-3-haiku-20240307` - Fastest, most affordable
  </Step>

  <Step title="Test Connection">
    Open the AI panel and send a test message. You should receive a response from Claude.
  </Step>
</Steps>

## Configuration

### Provider Settings

* **Service**: `anthropic`
* **Base URL**: `https://api.anthropic.com` (default)
* **Authentication**: `x-api-key` header
* **API Version**: `2023-06-01` (automatically set)

### Custom Endpoint

To use a custom Anthropic endpoint (proxy, self-hosted):

1. Set **Base URL** to your endpoint (e.g., `https://your-proxy.com`)
2. Add any required headers in **Custom Headers**
3. Enable **Allow Private Hosts** if using localhost

## Model Selection

Glyph fetches the latest model list from Anthropic's `/v1/models` endpoint.

### Recommended Models

| Model                        | Use Case                    | Context Window | Max Output |
| ---------------------------- | --------------------------- | -------------- | ---------- |
| `claude-3-5-sonnet-20241022` | General purpose, coding     | 200K tokens    | 8K tokens  |
| `claude-3-opus-20240229`     | Complex reasoning, analysis | 200K tokens    | 4K tokens  |
| `claude-3-sonnet-20240229`   | Balanced tasks              | 200K tokens    | 4K tokens  |
| `claude-3-haiku-20240307`    | Fast, simple tasks          | 200K tokens    | 4K tokens  |

<Note>
  Claude models have a **200K token context window**, allowing you to attach large amounts of context from your notes.
</Note>

### Max Tokens Requirement

Anthropic requires the `max_tokens` parameter for all requests. Glyph automatically sets this to **2048 tokens** for Anthropic models.

If you need longer responses, this value is hardcoded in `src-tauri/src/ai_rig/runtime.rs:73`.

## Features

### Chat Mode

Conversational interaction without tools:

* Back-and-forth dialogue with Claude
* No file system access
* Faster responses
* Best for discussion and brainstorming

### Create Mode

Claude with workspace tools:

* **read\_file** - Read files from your space
* **search\_notes** - Search note content
* **list\_dir** - List directory contents
* Tool usage shown in timeline view
* Best for research, summarization, and analysis

### Context Attachment

Attach notes to leverage Claude's 200K context window:

* Attach files or folders via context menu
* Mention with `@filename` syntax
* Configure character budget (up to 250K chars)
* View token estimates before sending

## API Usage and Billing

Glyph makes direct API calls to Anthropic:

* You are billed by Anthropic based on token usage
* No additional fees from Glyph
* Check usage at [console.anthropic.com](https://console.anthropic.com)

### Cost Estimation

| Model             | Input (per 1M tokens) | Output (per 1M tokens) |
| ----------------- | --------------------- | ---------------------- |
| Claude 3.5 Sonnet | \$3.00                | \$15.00                |
| Claude 3 Opus     | \$15.00               | \$75.00                |
| Claude 3 Sonnet   | \$3.00                | \$15.00                |
| Claude 3 Haiku    | \$0.25                | \$1.25                 |

Use the context manifest to estimate costs before sending.

## Rate Limits

Anthropic enforces rate limits based on your usage tier:

* Limits vary by model and tier
* Check your tier at [console.anthropic.com](https://console.anthropic.com)
* Rate limit errors include retry-after headers

If you hit rate limits, Glyph displays the error. Wait before retrying.

## Troubleshooting

### "API key not set for this profile"

**Solution**: Add your Anthropic API key in Settings → AI.

### "model list failed (401)"

**Solution**: Your API key is invalid or expired. Generate a new key from Anthropic Console.

### "model list failed (403)"

**Solution**: Your API key doesn't have permission to list models. This is non-fatal; type the model ID manually.

### "model list failed (429)"

**Solution**: You've hit Anthropic's rate limit. Wait before retrying.

### Model list is empty

**Solution**: Type the model ID manually:

* `claude-3-5-sonnet-20241022`
* `claude-3-opus-20240229`
* `claude-3-haiku-20240307`

The model will work even if the list fetch failed.

### Responses are truncated

**Cause**: Anthropic's `max_tokens` parameter (default 2048) limits output length.

**Solution**: To increase output length, modify `src-tauri/src/ai_rig/runtime.rs:73` and rebuild:

```rust theme={null}
let max_tokens = if caps.requires_max_tokens {
    Some(4096)  // Increase from 2048
} else {
    None
};
```

### "context length exceeded"

**Cause**: Total tokens (context + conversation) exceeds 200K.

**Solution**: Reduce attached context or use a smaller character budget.

## Security Best Practices

<Warning>
  * Never commit `.glyph/app/ai_secrets.json` to version control
  * Rotate API keys if exposed
  * Use separate keys for different projects
  * Monitor usage in Anthropic Console
</Warning>

## Claude-Specific Tips

### Extended Thinking

Claude models excel at reasoning tasks. In create mode, Claude can:

* Read multiple files to synthesize information
* Search your notes for relevant context
* Chain multiple tool calls to answer complex questions

### System Prompts

Claude respects system prompts well. In create mode, Glyph adds:

```
Tool discipline for this run: use the minimum number of tool calls needed.
Prefer at most 1-2 search/list calls before answering.
If a tool returns usable evidence, stop searching and summarize.
```

This reduces unnecessary tool usage and improves response speed.

### Markdown and Code

Claude generates well-formatted markdown and code. Responses render beautifully in Glyph's AI panel.

## Next Steps

<CardGroup cols={2}>
  <Card title="Chat Modes" icon="messages" href="/ai/chat-modes">
    Learn about chat vs create modes
  </Card>

  <Card title="Context Management" icon="folder" href="/ai/context-management">
    Attach notes to leverage Claude's 200K context
  </Card>

  <Card title="OpenAI" icon="brain" href="/ai/providers/openai">
    Compare with OpenAI GPT models
  </Card>

  <Card title="Profiles" icon="user-group" href="/ai/profiles">
    Manage multiple AI profiles
  </Card>
</CardGroup>