Stream response

Overview

The stream response endpoint returns chat completions in real time using Server-Sent Events (SSE). Instead of waiting for the full response, you receive chunks as they are generated.

Endpoint

POST/v1/chat/completions

Streaming uses the same endpoint as regular chat completions. Set stream: true in the request body.

Authentication

This endpoint requires a Bearer token in the Authorization header:

Authorization: Bearer sk_unii_your_api_key_here

Authorization: Bearer sk_unii_your_api_key_here

Request body

The request body accepts the same parameters as Chat completions, with stream set to true:

Parameter	Type	Description
`model`*	string	The model to use. Use `unii-latest` for the current default.
`messages`*	array[object]	An array of message objects with `role` and `content`.
`stream`*	boolean	Must be set to `true` to enable streaming. This is the only difference from the regular chat completions endpoint.
`temperature`	number	Controls randomness. Range: `0` to `2`. Default: `1`.
`max_tokens`	integer	Maximum number of tokens to generate.

Request examples

curl -X POST https://api.unii.dev/v1/chat/completions \
-H "Authorization: Bearer $UNII_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "model": "unii-latest",
  "messages": [
    { "role": "user", "content": "Explain RAG in detail." }
  ],
  "stream": true
}'

curl -X POST https://api.unii.dev/v1/chat/completions \-H "Authorization: Bearer $UNII_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "unii-latest",  "messages": [    { "role": "user", "content": "Explain RAG in detail." }  ],  "stream": true}'

Response format

The response is an SSE stream. Each event is prefixed with data: and contains a JSON chunk:

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"RAG"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":" stands"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"RAG"},"finish_reason":null}]}data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":" stands"},"finish_reason":null}]}data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}data: [DONE]

Chunk structure

Each chunk contains:

chunk.json

{
"id": "chatcmpl-abc123",
"object": "chat.completion.chunk",
"choices": [
  {
    "index": 0,
    "delta": {
      "role": "assistant",
      "content": "RAG"
    },
    "finish_reason": null
  }
]
}

chunk.json

{"id": "chatcmpl-abc123","object": "chat.completion.chunk","choices": [  {    "index": 0,    "delta": {      "role": "assistant",      "content": "RAG"    },    "finish_reason": null  }]}

Field	Type	Description
`id`	string	The completion ID
`object`	string	Always `chat.completion.chunk`
`choices[index].delta`	object	The new content fragment
`choices[index].finish_reason`	string	`null` during streaming, `"stop"` on the last chunk

End of stream

The stream ends with a final chunk containing finish_reason: "stop" followed by data: [DONE].

Note

The final chunk may have an empty delta object. Use finish_reason to detect the end of the stream.

Errors

If an error occurs during streaming, the SDK throws an exception. With raw HTTP, the connection closes with an error status code.

Important

The streaming protocol details shown on this page are illustrative. Verify the actual event format and chunk structure against the latest SDK documentation before using them in production.

Next steps

Read the Chat completions endpoint for non-streaming requests.
Try the Streaming guide for implementation examples.
See the API Overview for a full list of endpoints.

Stream response

OverviewCopy Anchor Link

EndpointCopy Anchor Link

AuthenticationCopy Anchor Link

Request bodyCopy Anchor Link

Request examplesCopy Anchor Link

Response formatCopy Anchor Link

Chunk structureCopy Anchor Link

End of streamCopy Anchor Link

ErrorsCopy Anchor Link

Next stepsCopy Anchor Link

On this page