Last Update: 2025/3/26

SenseFlow Chat API

The SenseFlow Chat API allows you to build conversational AI applications with context awareness. This is suitable for chatbots, virtual assistants, and other applications requiring conversation history.

Endpoints

Chat Message

POST https://platform.llmprovider.ai/v1/agent/chat-messages

Request Headers

Header	Value
Authorization	Bearer YOUR_API_KEY
Content-Type	application/json

Request Body

Parameter	Type	Description
model	string	agent name
query	string	User input/question content
inputs	object	(Optional) Key-value pairs of variables defined in the app
files	array	(Optional) Array of file objects for multimodal interactions
response_mode	string	Response mode: `streaming` (recommended) or `blocking`
conversation_id	string	(Optional) Conversation ID for continuing previous chats
user	string	Unique identifier for the end user

Files Object Structure

Field	Type	Description
type	string	File type (`image`、`document`, `audio`, `video`, `custom`)
transfer_method	string	`remote_url` or `local_file`
url	string	File URL (when transfer_method is `remote_url`)
upload_file_id	string	File ID (when transfer_method is `local_file`)

Response

The response varies based on the response_mode:

For blocking mode: Returns a ChatResponse object
For streaming mode: Returns a stream of ChunkChatResponse objects

ChatResponse Structure

Field	Type	Description
id	string	Message ID
conversation_id	string	Conversation ID
answer	string	Complete response content
message_files	array	Array of message files (for assistant responses)
metadata	object	Metadata information
created_at	integer	Message creation timestamp

Example Response (Blocking Mode)

{
  "id": "msg_12345",
  "conversation_id": "conv_12345",
  "answer": "Based on the image, I can see...",
  "message_files": [],
  "metadata": {},
  "created_at": 1705569239
}

Streaming Response Events

Each streaming chunk starts with data: and chunks are separated by \n\n. The events follow the same structure as the Completion API.

Streaming Response Events

Each streaming chunk starts with data: and chunks are separated by \n\n. Example:

data: {"event": "message", "task_id": "900bbd43-dc0b-4383-a372-aa6e6c414227", "message_id": "663c5084", "answer": "Hi", "created_at": 1705398420}\n\n

Different event types in the stream:

Event Type	Description	Fields
message	text chunk from LLM	task_id, message_id, answer, created_at
message_file	a new file has created by tool	id, type, belongs_to, url, conversation_id
message_end	receiving this event means streaming has ended	task_id, message_id, metadata, usage, retriever_resources
tts_message	Speech synthesis chunk	task_id, message_id, audio (base64), created_at
tts_message_end	Speech synthesis completion	task_id, message_id, audio (empty), created_at
message_replace	Content moderation replacement	task_id, message_id, answer, created_at
workflow_started	workflow starts execution	task_id, workflow_run_id, event, data(id, workflow_id, sequence_number, created_at)
node_started	node execution started	task_id, workflow_run_id, event, data(id, node_id, node_type, title, index, predecessor_node_id, inputs, created_at)
node_finished	node execution ended	task_id, workflow_run_id, event, data(id, node_id, node_type, title, index, predecessor_node_id, inputs, created_at)
workflow_finished	workflow execution ended	task_id, workflow_run_id, event, data(id, workflow_id, status, outputs, error, elapsed_time, total_tokens, total_steps, created_at, finished_at)
error	Stream error event	task_id, message_id, status, code, message
ping	Keep-alive ping (every 10s)	-

Example Request

Shell
Python
Node.js

curl -X POST 'https://platform.llmprovider.ai/v1/agent/chat-messages' \
--header 'Authorization: Bearer $YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data-raw '{
    "model": ""
    "query": "What can you tell me about this image?",
    "files": [
        {
            "type": "image",
            "transfer_method": "remote_url",
            "url": "https://example.com/image.jpg"
        }
    ],
    "response_mode": "streaming",
    "conversation_id": "conv_12345",
    "user": "abc-123"
}'

import requests
import json

api_key = 'YOUR_API_KEY'
url = 'https://platform.llmprovider.ai/v1/agent/chat-messages'

headers = {
    'Authorization': f'Bearer {api_key}',
    'Content-Type': 'application/json'
}

data = {
    'agent': ''
    'query': 'What can you tell me about this image?',
    'files': [{
        'type': 'image',
        'transfer_method': 'remote_url',
        'url': 'https://example.com/image.jpg'
    }],
    'response_mode': 'streaming',
    'conversation_id': 'conv_12345',
    'user': 'abc-123'
}

response = requests.post(url, headers=headers, json=data)
print(response.json())

const axios = require('axios');

const apiKey = 'YOUR_API_KEY';
const url = 'https://platform.llmprovider.ai/v1/agent/chat-messages';

const data = {
    model: ''
    query: 'What can you tell me about this image?',
    files: [{
        type: 'image',
        transfer_method: 'remote_url',
        url: 'https://example.com/image.jpg'
    }],
    response_mode: 'streaming',
    conversation_id: 'conv_12345',
    user: 'abc-123'
};

const headers = {
    'Authorization': `Bearer ${apiKey}`,
    'Content-Type': 'application/json'
};

axios.post(url, data, {headers})
    .then(response => console.log(response.data))
    .catch(error => console.error(error));

Upload File

POST https://platform.llmprovider.ai/v1/agent/files/upload

Upload files (currently only supports images) for use in multimodal interactions. Supports png, jpg, jpeg, webp, and gif formats.

Request Headers

Header	Value
Authorization	Bearer YOUR_API_KEY
Content-Type	multipart/form-data

Request Body Parameters

Parameter	Type	Description
model	string	agent name
file	file	The file to upload
user	string	User identifier (must match the message API user ID)

Response

Field	Type	Description
id	uuid	File ID
name	string	File name
size	integer	File size in bytes
extension	string	File extension
mime_type	string	File MIME type
created_by	uuid	Uploader ID
created_at	timestamp	Upload timestamp

Example Response

{
  "id": "72fa9618-8f89-4a37-9b33-7e1178a24a67",
  "name": "example.png",
  "size": 1024,
  "extension": "png",
  "mime_type": "image/png",
  "created_by": 123,
  "created_at": 1577836800
}

Example Request

Shell
Python
Node.js

curl -X POST 'https://api.dify.ai/v1/agent/files/upload' \
--header 'Authorization: Bearer {api_key}' \
--form 'file=@localfile;type=image/[png|jpeg|jpg|webp|gif] \
--form 'model='
--form 'user=abc-123'

import requests
import json

api_key = 'YOUR_API_KEY'
url = 'https://api.dify.ai/v1/agent/files/upload'

headers = {
    'Authorization': f'Bearer {api_key}',
    'Content-Type': 'application/json'
}

files = {
    'model': ('')
    'file': ('image.jpg', open('localfile', 'rb'), 'image/jpeg'),
    'user': (None, 'abc-123')
}

response = requests.post(url, headers=headers, files=files)
print(response.json())

const FormData = require('form-data');
const fs = require('fs');
const axios = require('axios');

const apiKey = 'YOUR_API_KEY';
const url = 'https://platform.llmprovider.ai/v1/agent/files/upload';

const form = new FormData();
form.append('file', fs.createReadStream('localfile'));
form.append('user', 'abc-123');
form.append('model', '');

const headers = {
    ...form.getHeaders(),
    'Authorization': `Bearer ${apiKey}`
};

axios.post(url, form, {headers})
    .then(response => console.log(response.data))
    .catch(error => console.error(error));

Stop Response

POST https://platform.llmprovider.ai/v1/agent/chat-messages/:task_id/stop

Stop a streaming response. Only available for streaming mode.

Request Headers

Header	Value
Authorization	Bearer YOUR_API_KEY
Content-Type	application/json

Path Parameters

Parameter	Type	Description
task_id	string	Task ID obtained from the streaming response

Request Body Parameters

Parameter	Type	Required	Description
user	string	Yes	User identifier (must match the message API user ID)
model	string	model name

Response

Field	Type	Description
result	string	upload result

Example Response

{
  "result": "success"
}

Example Request

Shell
Python
nodejs

curl -X POST 'https://platform.llmprovider.ai/v1/agent/chat-messages/task_123/stop' \
--header 'Authorization: Bearer $YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data-raw '{
        "agent": "",
        "user": "abc-123"
}'

import requests
import json

api_key = 'YOUR_API_KEY'
url = 'https://platform.llmprovider.ai/v1/agent/chat-messages/task_123/stop'

headers = {
        'Authorization': f'Bearer {api_key}',
        'Content-Type': 'application/json'
}

data = {
        'model': '',
        'user': 'abc-123'
}

response = requests.post(url, headers=headers, json=data)
print(response.json())

const axios = require('axios');

const apiKey = 'YOUR_API_KEY';
const url = 'https://platform.llmprovider.ai/v1/agent/chat-messages/task_123/stop';

const data = {
    model: ''
    user: 'abc-123'
};

const headers = {
    'Authorization': `Bearer ${apiKey}`,
    'Content-Type': 'application/json'
};

axios.post(url, data, {headers})
    .then(response => console.log(response.data))
    .catch(error => console.error(error));

Endpoints​

Chat Message​

Request Headers​

Request Body​

Files Object Structure​

Response​

ChatResponse Structure​

Example Response (Blocking Mode)​

Streaming Response Events​

Streaming Response Events​

Example Request​

Upload File​

Request Headers​

Request Body Parameters​

Response​

Example Response​

Example Request​

Stop Response​

Request Headers​

Path Parameters​

Request Body Parameters​

Response​

Example Response​

Example Request​

Endpoints

Chat Message

Request Headers

Request Body

Files Object Structure

Response

ChatResponse Structure

Example Response (Blocking Mode)

Streaming Response Events

Streaming Response Events

Example Request

Upload File

Request Headers

Request Body Parameters

Response

Example Response

Example Request

Stop Response

Request Headers

Path Parameters

Request Body Parameters

Response

Example Response

Example Request