# Model

AXAR AI supports multiple providers to interact with different language models. We can specify the provider and model identifier using the `@model` annotation when defining our agent. 

<details>

<summary>Setting up provider API keys</summary>

To use a specific provider, set up the required API key as an environment variable. Each provider has a specific environment variable name. For example, if you're using **Anthropic**, you can set the API key like this:

```bash
export ANTHROPIC_API_KEY="sk-proj-YOUR-API-KEY"
```

In development, you can simplify API key management by using the `dotenv` package along with a `.env` file. Here’s how to set it up:

1. **Install the `dotenv` package**\
   Use npm or yarn to install the `dotenv` package:

   ```bash
   npm install dotenv
   # or
   yarn add dotenv
   ```
2. **Create a `.env` file**\
   In the root directory of your project, create a `.env` file and add your API keys:

   ```bash
   ANTHROPIC_API_KEY=sk-proj-YOUR-API-KEY
   ```
3. **Load the `.env` file in your code**\
   Import and configure `dotenv` at the entry point of your application (e.g., `index.ts`):

   ```typescript
   import * as dotenv from 'dotenv';
   // Load environment variables from .env file
   dotenv.config();
   ```

</details>

### OpenAI

#### Supported models

<table data-full-width="false"><thead><tr><th width="307">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>gpt-4o</code></td><td><code>@model('openai:gpt-4o')</code></td></tr><tr><td><code>gpt-4o-mini</code></td><td><code>@model('openai:gpt-4o-mini')</code></td></tr><tr><td><code>gpt-4-turbo</code></td><td><code>@model('openai:gpt-4-turbo')</code></td></tr><tr><td><code>gpt-4</code></td><td><code>@model('openai:gpt-4')</code></td></tr><tr><td><code>o1</code></td><td><code>@model('openai:o1')</code></td></tr><tr><td><code>o1-mini</code></td><td><code>@model('openai:o1-mini')</code></td></tr></tbody></table>

{% hint style="info" %}
The API key for **OpenAI** is read from the `OPENAI_API_KEY` environment variable.
{% endhint %}

### Anthropic

#### Supported models

<table data-full-width="false"><thead><tr><th width="309">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>claude-3-5-sonnet-20241022</code></td><td><code>@model('anthropic:claude-3-5-sonnet-20241022')</code></td></tr><tr><td><code>claude-3-5-sonnet-20240620</code></td><td><code>@model('anthropic:claude-3-5-sonnet-20240620')</code></td></tr><tr><td><code>claude-3-5-haiku-20241022</code></td><td><code>@model('anthropic:claude-3-5-haiku-20241022')</code></td></tr></tbody></table>

{% hint style="info" %}
The API key for **Anthropic** is read from the `ANTHROPIC_API_KEY` environment variable.
{% endhint %}

### Google

#### Prerequisites

```bash
npm install @ai-sdk/google
```

{% hint style="info" %}
The API key for **Google** is read from the `GOOGLE_GENERATIVE_AI_API_KEY` env variable.
{% endhint %}

#### Supported models

<table data-full-width="false"><thead><tr><th width="315">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>gemini-2.0-flash-exp</code></td><td><code>@model('google:gemini-2.0-flash-exp')</code></td></tr><tr><td><code>gemini-1.5-flash</code></td><td><code>@model('google:gemini-1.5-flash')</code></td></tr><tr><td><code>gemini-1.5-pro</code></td><td><code>@model('google:gemini-1.5-pro')</code></td></tr></tbody></table>

### DeepSeek

#### Prerequisites

```bash
npm install @ai-sdk/deepseek
```

{% hint style="info" %}
The API key for **DeepSeek** is read from the `DEEPSEEK_API_KEY` env var.
{% endhint %}

#### Supported models

<table data-full-width="false"><thead><tr><th width="317">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>deepseek-chat</code></td><td><code>@model('deepseek:deepseek-chat')</code></td></tr></tbody></table>

### Cerebras

#### Prerequisites

```bash
npm install @ai-sdk/cerebras
```

{% hint style="info" %}
The API key for **Cerebras** is read from the`CEREBRAS_API_KEY` environment variable.
{% endhint %}

#### Supported models

<table data-full-width="false"><thead><tr><th width="318">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>llama3.1-8b</code></td><td><code>@model('</code>cerebras<code>:llama3.1-8b')</code></td></tr><tr><td><code>llama3.1-70b</code></td><td><code>@model('</code>cerebras<code>:llama3.1-70b')</code></td></tr><tr><td><code>llama3.3-70b</code></td><td><code>@model('</code>cerebras<code>:llama3.3-70b')</code></td></tr></tbody></table>

### Groq

#### Prerequisites

```bash
npm install @ai-sdk/groq
```

{% hint style="info" %}
The API key for **Groq** is read from the *`GROQ_API_KEY`* environment variable.
{% endhint %}

#### Supported models

<table data-full-width="false"><thead><tr><th width="321">Model</th><th>Usage</th></tr></thead><tbody><tr><td><code>llama-3.3-70b-versatile</code></td><td><code>@model('groq:llama-3.3-70b-versatile')</code></td></tr><tr><td><code>llama-3.1-8b-instant</code></td><td><code>@model('groq:llama-3.1-8b-instant')</code></td></tr><tr><td><code>mixtral-8x7b-32768</code></td><td><code>@model('groq:mixtral-8x7b-32768')</code></td></tr><tr><td><code>gemma2-9b-it</code></td><td><code>@model('groq:gemma2-9b-it')</code></td></tr></tbody></table>

{% hint style="success" %}
We can support all models and providers found here, if there's one missing or not working let us know: <https://sdk.vercel.ai/docs/foundations/providers-and-models>
{% endhint %}


---

# Agent Instructions: Querying This Documentation

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

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://axar-ai.gitbook.io/axar/basics/model.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.