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

# Models

> Available models and capabilities

## Overview

Anannas provides access to models from multiple providers through a unified API. Models are identified by `provider/model-name` format.

## Model Endpoint

Query available models:

```
GET https://api.anannas.ai/v1/models
```

### Response Format

```json theme={null}
{
  "data": [
    {
      "id": "openai/gpt-5-mini",
      "name": "GPT-5 Mini",
      "provider": "openai",
      "description": "Most capable GPT-5 Mini model",
      "context_length": 8192,
      "pricing": {
        "prompt": "0.00003",
        "completion": "0.00006",
        "request": "0"
      },
      "capabilities": [
        "text",
        "tool_calling",
        "json_mode"
      ]
    }
  ]
}
```

## Model Object Schema

| Field            | Type       | Description                                       |
| ---------------- | ---------- | ------------------------------------------------- |
| `id`             | `string`   | Model identifier (e.g., `openai/gpt-5-mini`)      |
| `name`           | `string`   | Human-readable name                               |
| `provider`       | `string`   | Provider identifier (`openai`, `anthropic`, etc.) |
| `description`    | `string`   | Model description                                 |
| `context_length` | `number`   | Maximum context window in tokens                  |
| `pricing`        | `object`   | Cost structure (see below)                        |
| `capabilities`   | `string[]` | Supported features                                |

## Pricing Object

```typescript theme={null}
type Pricing = {
  prompt: string;      // Cost per input token (USD)
  completion: string; // Cost per output token (USD)
  request?: string;   // Fixed cost per request (USD)
};
```

Prices are returned as strings to avoid floating-point precision issues.

## Model Format

Models use the format `provider/model-name`:

* `openai/gpt-5-mini`
* `openai/gpt-3.5-turbo`
* `anthropic/claude-3-sonnet`
* `anthropic/claude-3-opus`
* `x-ai/grok-beta`
* `groq/llama-3-70b`

## Capabilities

<Card title="View Model Capabilities" icon="book">
  For complete capability listings by model, visit [anannas.ai/models](https://anannas.ai/models) to see which features each model supports.
</Card>

Models may support:

* `text`: Text generation
* `tool_calling`: Function/tool execution
* `json_mode`: Structured JSON output
* `streaming`: Server-Sent Events
* `multimodal`: Image/audio inputs
* `reasoning`: Extended reasoning (o1, o3, etc.)
* `audio_output`: Audio generation
* `image_output`: Image generation

## Provider-Specific Models

<Card title="View All Models" icon="book">
  For complete, up-to-date model listings, pricing, capabilities, and parameter support, visit [anannas.ai/models](https://anannas.ai/models).
</Card>

### OpenAI

* `openai/gpt-5-mini`
* `openai/gpt-4-turbo`
* `openai/gpt-3.5-turbo`
* `openai/o1-preview`
* `openai/o1-mini`
* `openai/o3-mini`

### Anthropic

* `anthropic/claude-3-opus`
* `anthropic/claude-3-sonnet`
* `anthropic/claude-3-haiku`
* `anthropic/claude-sonnet-4-5`

### xAI (Grok)

* `x-ai/grok-beta`
* `x-ai/grok-2`

### Other Providers

Models from Groq, TogetherAI, DeepInfra, Fireworks, Nebius, and others are available. Query `/v1/models` for the complete list.

## Model Selection

### Default Model

If `model` is omitted, Anannas uses your account's default model. Set defaults in the dashboard.

### Model Routing

Use `provider` preferences to control model selection:

```json theme={null}
{
    "model": "openai/gpt-5-mini",
  "provider": {
    "order": ["openai", "anthropic"],
    "sort": "price"
  }
}
```

### Fallbacks

Specify fallback models:

```json theme={null}
{
    "model": "openai/gpt-5-mini",
  "fallbacks": [
    "anthropic/claude-3-sonnet",
    "openai/gpt-3.5-turbo"
  ]
}
```

## Parameter Support

<Card title="Check Parameter Support" icon="book">
  For detailed parameter support by model, visit [anannas.ai/models](https://anannas.ai/models) to see which parameters each model supports.
</Card>

Not all models support all parameters. Common support:

* **Temperature**: Most models
* **Top P**: Most models
* **Max Tokens**: All models
* **Tool Calling**: OpenAI, Anthropic (Claude 3+)
* **JSON Mode**: OpenAI GPT-5 Mini, Claude 3+
* **Streaming**: All models
* **Reasoning**: o1, o3, Claude Sonnet 4.5

## Context Windows

<Card title="View Context Windows" icon="book">
  For accurate context window sizes for all models, check [anannas.ai/models](https://anannas.ai/models).
</Card>

Context window sizes vary by model:

* GPT-5 Mini: 128,000 tokens
* GPT-4 Turbo: 128,000 tokens
* Claude 3 Opus: 200,000 tokens
* Claude 3 Sonnet: 200,000 tokens
* o1: 128,000 tokens

Exceeding context limits returns an error.

## Rate Limits

Rate limits are tier-based and vary by model. Check your tier in the dashboard. Limits apply per API key.

## Pricing

<Card title="View Current Pricing" icon="book">
  For up-to-date pricing for all models, visit [anannas.ai/models](https://anannas.ai/models). Pricing is updated in real-time.
</Card>

Pricing is per-token and varies by model. Check the `/v1/models` endpoint for current rates. Prices are in USD.

Example pricing (approximate):

* GPT-5 Mini: $0.0015/1K input, $0.002/1K output
* GPT-3.5 Turbo: $0.0015/1K input, $0.002/1K output
* Claude 3 Sonnet: $0.003/1K input, $0.015/1K output

## Model Updates

Models are updated automatically. Model identifiers remain stable, but underlying model versions may change. Check model descriptions for version information.

## Deprecated Models

Deprecated models continue to work but may be removed in future versions. Migrate to supported alternatives.

## See Also

* [API Overview](/API/Overview) - Request format
* [Parameters](/API/Parameters) - Parameter support
* [Quickstart](/quickstart) - Getting started