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

# Cerebras

> Cerebras Systems achieves the industry's lowest inference latency using wafer-scale chip architecture, delivering token generation speeds orders of magnitude faster than GPU-based providers.

Cerebras offers 2 models through Lava's AI Gateway, supporting Chat Completions. Authentication uses `Authorization: Bearer`. See the [Cerebras API docs](https://docs.cerebras.ai) for provider-specific parameters.

> [!NOTE]
>
> Supports both **managed** (Lava's API keys) and **unmanaged** (bring your own credentials) mode.

## Quick Start

```typescript theme={null}
const response = await fetch('https://api.lava.so/v1/forward?u=https%3A%2F%2Fapi.cerebras.ai%2Fv1%2Fchat%2Fcompletions', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    Authorization: `Bearer ${forwardToken}`,
  },
  body: JSON.stringify({
    model: 'gpt-oss-120b',
    messages: [{ role: "user", content: "Hello!" }],
  }),
});
```

## Chat Completions

**Target URL:** `https://api.cerebras.ai/v1/chat/completions`

| | |
| - | - |
| **Content Type** | `application/json` |
| **Streaming** | Yes (set `stream: true` in request body) |

| Model | Input / 1M tokens | Output / 1M tokens |
| - | - | - |
| qwen-3.8-27b | \$0.99 | \$1.49 |
| gpt-oss-120b | \$0.35 | \$0.75 |

## Next Steps

> [card] [All Providers](https://lava.so/docs/gateway/supported-providers.md)
>
> Browse all supported AI providers

> [card] [Forward Proxy](https://lava.so/docs/gateway/forward-proxy.md)
>
> Learn how to construct proxy URLs and authenticate requests


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.