RAG 问答

POST

/v1/knowledge/chat

基于知识库执行语义检索，将命中片段作为上下文调用 LLM 生成回答。

Request

Body Params application/jsonRequired

Examples

Responses

🟢200

application/json

成功（OpenAI 兼容格式；stream=true 时为 text/event-stream）

Bodyapplication/json

🟠401

Request Example

Shell

JavaScript

Java

Swift

curl --location 'https://api.aboai.ai/v1/knowledge/chat' \
--header 'Content-Type: application/json' \
--data '{
    "question": "什么是新餐饮",
    "kid": "AbCd123456",
    "topK": 5,
    "stream": false,
    "model": "",
    "similarityThreshold": 0.55,
    "returnSources": false
}'

Response Example

200 - 非流式响应成功示例

{
  "id": "resp_abc123",
  "provider": "openai",
  "model": "gpt-4o",
  "content": "根据产品手册，自签收之日起 7 天内支持无理由退换货，退换货需保证商品及包装完好。",
  "thinking": "思考过程",
  "tool_calls": [
    {
      "id": "call_abc123",
      "type": "function",
      "function": {
        "name": "get_weather",
        "arguments": "{\"city\":\"上海\",\"unit\":\"celsius\"}"
      }
    }
  ],
  "finish_reason": "stop",
  "usage": {
    "input_tokens": 1250,
    "output_tokens": 85,
    "cached_input_tokens": 0,
    "reasoning_tokens": 0
  },
  "raw_usage": {
    "prompt_tokens": 1250,
    "completion_tokens": 85,
    "total_tokens": 1335
  }
}

Modified at 2026-08-14 08:56:00

语义检索

AboAI API 对接Claude Code