[BUG] chatflows adding multiple times (per every request) in to the ChatFlowPool using Embed website or API

[IacsonNicolas]

Describe the bug
Every time i make a request from JS, Postman or even embed in websites the request takes a considerable amount of time in retriving a answer (if it does) and takes around a x1000 of tokens in the open IA embedding model, triying to find the error i found that for every request is added a chatflow in to the ChatflowPool

To Reproduce
Steps to reproduce the behavior:

1. Create a flow with embedding a model of Open IA
2. Make curl or use postman to call the flowise api
3. See chatflow added multiple times in console and billing in Open IA

Expected behavior
The api should not add the chatflow in the ChatflowPool every time a request is created from the api or the embed (i think the problem comes from there)
The amount of tokens used per request should be similar to the amount of tokens used in the chat integrated in flowise and the responses should be quicker

Screenshots

Setup

• Installation [npx flowise start]
• Flowise Version [2.1.2]
• OS: [Windows]
• Browser [Firefox]

[IacsonNicolas]

it also happens in docker and others versions of flowise

this is the flow i used, hope it helps
Chatflow.json

And this is a demostration with ollama
https://www.youtube.com/watch?v=PVvGOUJEeW0

[HenryHengZJ]

ah yes, when you use In Memory Store, every time when the question is asked, it will do the embeddings everytime. recommend to use others.. we should have some description around mentioning that