> For the complete documentation index, see [llms.txt](https://missiveapp.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://missiveapp.com/docs/ai/setup/anthropic.md).

# Anthropic Claude

Connect your Anthropic account to use Claude models in Missive.

Connect your Anthropic account to use Claude models for the assistant, prompts, and AI rules.

{% hint style="warning" %}
Anthropic API access requires prepaid credits or a billing plan. A **Claude Pro subscription** ($20/month) is not the same as API access. You need a separate API key from the Anthropic Console.
{% endhint %}

## Setup

{% stepper %}
{% step %}
**Get your Anthropic API key**

Create an account at [console.anthropic.com](https://console.anthropic.com/) if you don't have one.

Go to **API Keys** and click **Create Key**. Copy the key somewhere safe. You won't be able to see it again.

Make sure you have [billing set up](https://console.anthropic.com/settings/billing) with prepaid credits or a payment method.
{% endstep %}

{% step %}
**Add the integration in Missive**

In Missive, go to **Settings** > **Integrations** > click **Add integration** > search for **Anthropic** and click **Add to Missive**.

Enter your **API key** and click **Add**.
{% endstep %}

{% step %}
**Share with your team (optional)**

To let your team use AI features, share the integration:

Go to **Settings** > **Integrations** > **Anthropic** and click **Share integration** at the top.

{% hint style="warning" %}
Shared integrations mean everyone uses the same API key. All usage bills to your Anthropic account.
{% endhint %}
{% endstep %}

{% step %}
**Restrict available models (optional)**

By default, all models are available to your team. To limit which models they can choose in the Assistant:

Go to **Settings** > **Integrations** > **Anthropic** and scroll to **Models**.

Select the models you want to make available. Choosing **All models** automatically includes any future models. Deselecting specific models prevents your team from using them.
{% endstep %}
{% endstepper %}

## Available models <a href="#available-models" id="available-models"></a>

Missive supports every current Claude model. New models appear automatically as Anthropic releases them.

### Assistant models

| Model             | Best for                                              | Speed    | Quality |
| ----------------- | ----------------------------------------------------- | -------- | ------- |
| Claude Opus 5.5   | Complex reasoning, nuanced writing, detailed analysis | Moderate | Highest |
| Claude Sonnet 5.5 | Good balance of speed and quality, most daily tasks   | Fast     | High    |
| Claude Haiku 5.5  | Quick tasks, lowest cost                              | Fastest  | Good    |

For drafting customer-facing email replies, Claude Opus 5.5 gives the best results. Claude Sonnet 5.5 is a good all-rounder for most daily tasks.

Older models are still available under **Legacy** in the model picker.

### Rule models

AI actions in rules can use any of the models above. **Fast (synchronous)** models run inline, so the labels they apply can trigger other rules. **Prompt** conditions only support Fast models. See [AI Rules](/docs/advanced-features/rules/ai-rules.md#model-selection) for details.

| Model            | Type                                                                          |
| ---------------- | ----------------------------------------------------------------------------- |
| Claude Haiku 5.5 | Fast. Use in Prompt conditions or chained rules                               |
| All other models | Powerful (asynchronous). Better results, but labels won't trigger other rules |

## Data privacy

By default, Anthropic does **not** use API inputs or outputs to train models. Data submitted through the API is retained for 30 days for safety monitoring, then deleted.

Review your settings at [platform.claude.com/settings/privacy](https://platform.claude.com/settings/privacy):

* **Allow user feedback** controls whether feedback on model responses (including full prompts and responses) is sent to Anthropic.
* **Development Partner Program** is an opt-in program that shares data with Anthropic for model training. Do not join if you want to keep your data private.

## Cost and usage

Anthropic bills based on tokens processed (input and output). A token is roughly 3/4 of a word. Check [Anthropic's pricing page](https://platform.claude.com/docs/en/about-claude/pricing) for current rates.

Each request includes the conversation, your instructions, and the assistant's tool definitions, so reading a thread and drafting a reply typically uses 10,000 to 30,000 tokens. With Claude Sonnet 5.5, that's usually a few cents per request.

### Managing costs

* **Use the best model for manual tasks.** When you trigger AI yourself (drafting a reply, asking the assistant a question), the cost per interaction is negligible. Use Claude Opus 5.5 for the best results.
* **Use Haiku for AI rules.** Rules run automatically on every matching message, so costs add up. Claude Haiku 5.5 is fast and cheap for classification and routing.
* **Set spending limits** in your [Anthropic account settings](https://console.anthropic.com/settings/limits).
* **Monitor usage** in the [Anthropic Console](https://console.anthropic.com/settings/usage).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://missiveapp.com/docs/ai/setup/anthropic.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.