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

# Chat Completions

> Generate text using the OpenAI-compatible chat completions endpoint.

# Chat Completions

```
POST /v1/chat/completions
```

Generate text from a conversation. Supports streaming, vision, and structured output.

## Request

```bash theme={null}
curl https://inference.dria.co/v1/chat/completions \
  -H "Authorization: Bearer dkn_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.5:9b",
    "messages": [
      {"role": "user", "content": "explain quantum computing in one sentence"}
    ]
  }'
```

### Parameters

| Field             | Type    | Required | Description                              |
| ----------------- | ------- | -------- | ---------------------------------------- |
| `model`           | string  | yes      | Model ID (e.g., `qwen3.5:9b`)            |
| `messages`        | array   | yes      | Conversation messages                    |
| `max_tokens`      | integer | no       | Max tokens to generate (default: `2048`) |
| `temperature`     | float   | no       | Sampling temperature (default: `0.7`)    |
| `stream`          | boolean | no       | Enable SSE streaming (default: `false`)  |
| `timeout_secs`    | integer | no       | Timeout in seconds (default: `120`)      |
| `response_format` | object  | no       | Structured output schema                 |

### Message format

Each message has a `role` and `content`:

```json theme={null}
{"role": "system", "content": "You are a helpful assistant"}
{"role": "user", "content": "Hello"}
{"role": "assistant", "content": "Hi! How can I help?"}
{"role": "user", "content": "What is Rust?"}
```

### Vision (multimodal)

For vision models, `content` can be an array of parts:

```json theme={null}
{
  "role": "user",
  "content": [
    {"type": "text", "text": "Describe this image"},
    {"type": "image_url", "image_url": {"url": "data:image/jpeg;base64,..."}}
  ]
}
```

### Structured output

Use `response_format` to get JSON conforming to a schema:

```json theme={null}
{
  "model": "qwen3.5:9b",
  "messages": [{"role": "user", "content": "John Doe, john@example.com, 30"}],
  "response_format": {
    "type": "json_schema",
    "json_schema": {
      "name": "extract",
      "schema": {
        "type": "object",
        "properties": {
          "name": {"type": "string"},
          "email": {"type": "string"},
          "age": {"type": "integer"}
        },
        "required": ["name", "email", "age"]
      }
    }
  }
}
```

## Response

```json theme={null}
{
  "id": "gen-abc123",
  "model": "qwen3.5:9b",
  "choices": [
    {
      "message": {
        "content": "Quantum computing uses quantum bits..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 15,
    "completion_tokens": 42,
    "total_tokens": 57
  },
  "metadata": {
    "node_id": "node-xyz"
  }
}
```

## Streaming

Set `"stream": true` to receive Server-Sent Events:

```bash theme={null}
curl https://inference.dria.co/v1/chat/completions \
  -H "Authorization: Bearer dkn_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.5:9b",
    "messages": [{"role": "user", "content": "hello"}],
    "stream": true
  }'
```

Each event is a `data:` line with a JSON chunk:

```
data: {"choices":[{"delta":{"content":"Hello"}}]}
data: {"choices":[{"delta":{"content":"!"}}]}
data: {"choices":[{"delta":{"content":" How"}}]}
data: [DONE]
```

The stream ends with `data: [DONE]`.