POST
/
v1
/
memory
/
chat
/
completions
curl -X POST 'https://api.sid.ai/v1/memory/chat/completions?namespace=user123' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer <access_token>' \
--data-raw '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
}'

{
  "id": "chatcmpl-123",
  "object": "chat.completion",
  "created": 1677652288,
  "model": "gpt-3.5-turbo-0613",
  "system_fingerprint": "fp_44709d6fcb",
  "choices": [{
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "\n\nHello there, how may I assist you today?",
    },
    "finish_reason": "stop"
  }],
  "usage": {
    "prompt_tokens": 9,
    "completion_tokens": 12,
    "total_tokens": 21
  }
}

​
Body & Response

The format of the request and response bodies is exactly the same as for the OpenAI chat completions API. See here for more details.

​
Query Parameters

​
namespace
string
default:""

The namespace from which to query the memory and where conversations are persisted. Defaults to the global namespace.

curl -X POST 'https://api.sid.ai/v1/memory/chat/completions?namespace=user123' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer <access_token>' \
--data-raw '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
}'

{
  "id": "chatcmpl-123",
  "object": "chat.completion",
  "created": 1677652288,
  "model": "gpt-3.5-turbo-0613",
  "system_fingerprint": "fp_44709d6fcb",
  "choices": [{
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "\n\nHello there, how may I assist you today?",
    },
    "finish_reason": "stop"
  }],
  "usage": {
    "prompt_tokens": 9,
    "completion_tokens": 12,
    "total_tokens": 21
  }
}