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

# AI Models

> Choose from Claude, GPT-5, and Gemini models for iOS development. Frontier AI optimized for Swift and SwiftUI code generation.

## Paid Plan Models

Available on Start, Pro, Max, and Ultra plans:

<CardGroup cols={2}>
  <Card title="Claude Sonnet 4.5">
    **Default model**

    Best coding model with extended autonomous operation and context awareness. Superior agent capabilities with enhanced tool usage and extended thinking.

    * Context: 200K tokens
    * Max output: 64K tokens
  </Card>

  <Card title="Claude Opus 4.5">
    **Frontier reasoning**

    Frontier reasoning model for complex software engineering, agentic workflows, and long-horizon tasks. Superior multimodal capabilities.

    * Context: 200K tokens
    * Max output: 32K tokens
  </Card>

  <Card title="GPT-5.2">
    **Flagship coding**

    OpenAI's flagship model for coding and agentic tasks. Superior reasoning capabilities with extended context for complex workflows.

    * Context: 272K tokens
    * Max output: 128K tokens
  </Card>

  <Card title="Gemini 3 Pro">
    **Multimodal reasoning**

    Google's flagship frontier model for high-precision multimodal reasoning across text, image, video, audio, and code.

    * Context: 200K tokens
    * Max output: 65K tokens
  </Card>
</CardGroup>

## Free Plan Models

Available with your own API keys:

<CardGroup cols={3}>
  <Card title="Claude Haiku 4.5">
    **Fast and efficient**

    Near-frontier intelligence at 2x speed. Matches Sonnet 4 performance at one-third the cost.

    * Context: 200K tokens
    * Max output: 64K tokens
  </Card>

  <Card title="GPT-5">
    **Advanced reasoning**

    OpenAI's most advanced model for step-by-step reasoning, coding, and complex multi-turn tasks.

    * Context: 272K tokens
    * Max output: 128K tokens
  </Card>

  <Card title="Gemini 2.5 Flash">
    **Cost-efficient**

    Google's workhorse model for reasoning, coding, and scientific tasks. Built-in thinking capabilities.

    * Context: 1M tokens
    * Max output: 65K tokens
  </Card>
</CardGroup>

## Switching Models

1. Click the model selector in the chat input area
2. Choose your preferred model
3. Continue your conversation

## Model Availability

| Model             | Free Plan (API key)             | Paid Plans                      |
| ----------------- | ------------------------------- | ------------------------------- |
| Claude Haiku 4.5  | <Icon icon="check" size={16} /> | —                               |
| GPT-5             | <Icon icon="check" size={16} /> | —                               |
| Gemini 2.5 Flash  | <Icon icon="check" size={16} /> | —                               |
| Claude Sonnet 4.5 | —                               | <Icon icon="check" size={16} /> |
| Claude Opus 4.5   | —                               | <Icon icon="check" size={16} /> |
| GPT-5.2           | —                               | <Icon icon="check" size={16} /> |
| Gemini 3 Pro      | —                               | <Icon icon="check" size={16} /> |

## Free Plan Setup

On the Free plan, provide your own API keys:

* **Anthropic API Key**: For Claude Haiku 4.5
* **OpenAI API Key**: For GPT-5
* **Google API Key**: For Gemini 2.5 Flash

See [API Keys](/api-keys) for setup instructions.