Completion
POSThttps://knox.chat/v1/completions
Send a completion request to the selected model (in plain text format)
Request
This endpoint requires an object containing the following properties:
Headers
Name | Type | Required | Description |
---|---|---|---|
Authorization | String | Yes | Bearer authentication in the form Bearer token , where token is your authorization token. |
Request Body
Name | Type | Required | Description |
---|---|---|---|
model | String | Yes | The model ID to use. If not specified, user's default settings will be used. |
prompt | String | Yes | The text prompt to complete |
models | List of Strings | No | Alternative model list for routing override. |
provider | Object | No | Provider routing preferences. |
reasoning | Object | No | Configuration for model reasoning/thinking tokens |
usage | Object | No | Whether to include usage information in the response |
transforms | List of Strings | No | List of prompt transformations (Knox Chat only). |
stream | Boolean | No | Enable result streaming. Defaults to false |
max_tokens | Integer | No | Maximum number of tokens (range: [1, context_length)). |
temperature | Double | No | Sampling temperature (range: [0, 2]). |
seed | Integer | No | Seed for deterministic output. |
top_p | Double | No | Top-p sampling value (range: (0, 1]). |
top_k | Integer | No | Top-k sampling value (range: [1, Infinity)). |
frequency_penalty | Double | No | Frequency penalty (range: [-2, 2]). |
presence_penalty | Double | No | Presence penalty (range: [-2, 2]). |
repetition_penalty | Double | No | Repetition penalty (range: (0, 2]). |
logit_bias | Map of String to Double | No | Mapping from token IDs to bias values. |
top_logprobs | Integer | No | Number of top log probabilities to return. |
min_p | Double | No | Minimum probability threshold (range: [0, 1]). |
top_a | Double | No | Alternative top sampling parameter (range: [0, 1]). |
cURL Example
curl -X POST https://knox.chat/v1/completions \
-H "Authorization: Bearer <token>" \
-H "Content-Type: application/json" \
-d '{
"model": "model",
"prompt": "您好!"
}'
Response
Success Response (200)
{
"id": "id",
"choices": [
{
"text": "我是一个AI助手,有什么可以帮您?",
"index": 1,
"finish_reason": "stop"
}
]
}
Response Schema
Name | Type | Description |
---|---|---|
id | string | Unique identifier for the completion |
choices | array | Array of completion choices |
choices[].text | string | Generated text |
choices[].index | integer | Index of the choice in the array (e.g.: 0, 1, 2) |
choices[].finish_reason | string | Reason for completion (stop, length, content_filter) |