Chat API

Chat API allows you to integrate an AI consultant into your application. Standard request-response mode and SSE streaming for real-time display are supported.

POST/api/v1/chat

Send Message

Sends a message to AI chat and receives a complete response. For real-time display, use the SSE endpoint.

Request Parameters

Name
message
Type
string
Description
User message text (max 4000 characters)
Name
chat_id
Type
integer
Description
Existing chat ID to continue the conversation
Name
stream
Type
boolean
Description
Must be `false` for this endpoint

Response

Name
chat_id
Type
integer
Description
Chat ID (use to continue conversation)
Name
message_id
Type
integer
Description
Created message ID
Name
response
Type
string
Description
AI text response
Name
product_ids
Type
array
Description
Array of recommended product IDs

Request

curl -X POST "https://api.parsewise.ru/v1/chat" \
  -H "Authorization: Bearer pw_your_token" \
  -H "Content-Type: application/json" \
  -d '{
    "message": "What running shoes do you recommend?",
    "stream": false
  }'

Response

{
  "chat_id": 456,
  "message_id": 789,
  "response": "For running, I recommend these models...",
  "product_ids": [12, 45, 78]
}

POST/api/v1/chat/stream

SSE Streaming

Sends a message and receives a response in real-time via Server-Sent Events. Perfect for creating a ChatGPT-like UX.

Request Parameters

Name
message
Type
string
Description
User message text
Name
chat_id
Type
integer
Description
Existing chat ID to continue conversation

Event Types

Name
products
Type
object
Description
Sent first, contains `product_ids`
Name
text
Type
object
Description
Response text parts in `content` field
Name
done
Type
object
Description
Completion, contains `message_id` and `chat_id`
Name
error
Type
object
Description
Error, contains `message`

Request

import requests
import json

response = requests.post(
    "https://api.parsewise.ru/v1/chat/stream",
    headers={"Authorization": "Bearer pw_your_token"},
    json={"message": "Tell me about products"},
    stream=True
)

for line in response.iter_lines():
    if line:
        line = line.decode('utf-8')
        if line.startswith('data: '):
            data = json.loads(line[6:])
            
            if data['type'] == 'products':
                print(f"Products: {data['data']['product_ids']}")
            elif data['type'] == 'text':
                print(data['data']['content'], end='', flush=True)
            elif data['type'] == 'done':
                print(f"\nChat ID: {data['data']['chat_id']}")

SSE Events

data: {"type": "products", "data": {"product_ids": [12, 45]}}

data: {"type": "text", "data": {"content": "We have "}}

data: {"type": "text", "data": {"content": "great options..."}}

data: {"type": "done", "data": {"message_id": 789, "chat_id": 456}}

GET/api/v1/chat/{chat_id}/messages

Message History

Returns all messages in the specified chat.

URL Parameters

Name
chat_id
Type
integer
Description
Chat ID

Response

Array of messages:

Name
id
Type
integer
Description
Message ID
Name
content
Type
string
Description
Message text
Name
is_from_user
Type
boolean
Description
`true` if from user, `false` if from AI
Name
created_at
Type
string
Description
Creation time in ISO 8601

Request

curl "https://api.parsewise.ru/v1/chat/456/messages" \
  -H "Authorization: Bearer pw_your_token"

Response

[
  {
    "id": 1,
    "content": "What running shoes do you have?",
    "is_from_user": true,
    "created_at": "2024-01-15T10:30:00Z"
  },
  {
    "id": 2,
    "content": "I recommend checking out...",
    "is_from_user": false,
    "created_at": "2024-01-15T10:30:05Z"
  }
]