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

# ExuluDefaultProviders

> Pre-built ExuluProvider instances for Anthropic, Cerebras, Google Vertex, and OpenAI — import and register without writing provider config.

`ExuluDefaultProviders` is a nested object containing ready-made `ExuluProvider` instances for the most commonly used models. Each entry is a fully configured provider that you can pass directly to `ExuluApp.create()` or reference when building an agent's model list in the platform UI.

The accessor path matches the `provider` key in the docs: `ExuluDefaultProviders.<vendor>.<accessor>`.

## Providers

### Anthropic

| Accessor             | Model ID            | Context        |
| -------------------- | ------------------- | -------------- |
| `anthropic.opus4`    | `claude-opus-4-0`   | 200 000 tokens |
| `anthropic.sonnet4`  | `claude-sonnet-4-0` | 200 000 tokens |
| `anthropic.sonnet45` | `claude-sonnet-4-5` | 200 000 tokens |

### Cerebras

| Accessor              | Model ID        | Context        |
| --------------------- | --------------- | -------------- |
| `cerebras.gptOss120b` | `gpt-oss-120b`  | 128 000 tokens |
| `cerebras.llama38b`   | `llama3.1-8b`   | 32 000 tokens  |
| `cerebras.llama3370b` | `llama-3.3-70b` | 32 000 tokens  |

Cerebras providers require a Cerebras API key configured as the provider credential in the platform. The `provider` field is `"custom"`.

### Google Vertex

| Accessor                     | Model ID                                   | Context          |
| ---------------------------- | ------------------------------------------ | ---------------- |
| `google.vertexGemini25Flash` | `gemini-2.5-flash`                         | 1 048 576 tokens |
| `google.vertexGemini25Pro`   | `gemini-2.5-pro`                           | 1 048 576 tokens |
| `google.vertexGemini3Pro`    | `gemini-3-pro-preview`                     | 1 048 576 tokens |
| `google.vertexLlamaScout4`   | `meta/llama-4-scout-17b-16e-instruct-maas` | 1 048 576 tokens |

Vertex providers require a Google Cloud service-account JSON as the provider credential. See the authentication note in the platform's provider settings.

### OpenAI

| Accessor           | Model ID       | Context          |
| ------------------ | -------------- | ---------------- |
| `openai.gpt5Mini`  | `gpt-5-mini`   | 400 000 tokens   |
| `openai.gpt5`      | `gpt-5`        | 400 000 tokens   |
| `openai.gpt5pro`   | `gpt-5-pro`    | 400 000 tokens   |
| `openai.gpt5Codex` | `gpt-5-codex`  | 400 000 tokens   |
| `openai.gpt5Nano`  | `gpt-5-nano`   | 400 000 tokens   |
| `openai.gpt41`     | `gpt-4.1`      | 1 047 576 tokens |
| `openai.gpt41Mini` | `gpt-4.1-mini` | 1 047 576 tokens |
| `openai.gpt4o`     | `gpt-4o`       | 128 000 tokens   |
| `openai.gpt4oMini` | `gpt-4o-mini`  | 128 000 tokens   |

## Example

```typescript theme={null}
import { ExuluApp, ExuluContext, ExuluDefaultProviders } from "@exulu/backend";

const app = new ExuluApp();

await app.create({
  providers: [
    ExuluDefaultProviders.anthropic.sonnet45,
    ExuluDefaultProviders.openai.gpt41,
    ExuluDefaultProviders.google.vertexGemini25Flash,
  ],
  contexts: [productContext],
  config: { /* ... */ },
});
```

## Notes

* Each default provider has a stable `id` (e.g. `claude_sonnet_4_5_provider`) that IMP stores in the database when a user selects it. Do not change the `id` without a migration.
* You can mix default providers with custom `ExuluProvider` instances in the same array.
* Model IDs in this table reflect the source at the time of the installed `@exulu/backend` version. Check the platform's model admin page for live availability.

## Related

* [ExuluProvider — introduction](/developers/core/exulu-provider/introduction): build and configure custom providers.
* [Administration / Models](/administration/models): manage model access and API keys in the platform.