Querying Flowise API from python doesn't use buffer memory?

[Hyperlooper]

Hello, I have a simple chain with buffer memory piped in to the conversational agent. This works properly using the embedded chat window in the webpage, but when I query the same chain via python using the API, the agent does not have any memory message to message. How do you query the agent via the API in a way that does preserve this memory state?

Working via UI

Does not work via this call:

def queryFlowise(payload):
response = requests.post(API_URL, json=payload)
print(response)
return response.json()

[HenryHengZJ]

@Hyperlooper
for buffer memory, you have to supply the list of chat messages as history in the API body - https://docs.flowiseai.com/integrations/memory/short-term-memory

alternative is to use memory nodes with sessionId like Redis, Zep, Upstash etc, and supply the sessionId in the body:
https://docs.flowiseai.com/integrations/memory/long-term-memory

[VincentEmmanuel]

can you elaborate on how to use buffer memory in the API?