Get started
1. Sign Up
Go to https://understand.tech/en and click on Get Started. Create your free account using your Google, GitHub, or email. After clicking Create Account, make sure to confirm your email address in your inbox.
2. Sign In
After creating your account, log in with your email, Google, or GitHub, and click Connect to continue.
And now… let’s get started!
3. Homepage overview 'Chat'
When you first log in, the homepage provides a central hub to access your AI assistants, manage your workspace, and customize your experience.
Use the Welcome Bot for any question that you have related to our platform!
EN / FR Toggle
Click this button to switch the platform’s language between English (EN) and French (FR).
Sun / Moon Toggle
Use this button to switch between Light mode (Sun) and Dark mode (Moon) for a comfortable viewing experience.
Chat Button
This allows you to interact with various AI assistants and large language models (LLMs).
All: Displays every AI model available in your account, including your own assistants, Understand Tech’s LLMs, and other public or private assistants.
LLM: Provides an overview of the large language models you can choose to chat with.
Private: Shows only the AI assistants you have personally created.
Public: Displays AI assistants created by others that are publicly available.
Shared: Displays solely the AI assistants that you have shared with someone
Team: Shows the AI assistants used and shared within your team
Workspace Button
User button
This section provides an overview of your usage:
Account: View your email, plan, usage stats, prompts, and available AI providers for chat and building assistants.
Plan: See your current subscription and upgrade or downgrade as needed.
Affiliation: Track referrals, earnings, and share your affiliate link. Set up payouts via Stripe.
Developer: Access your API key and explore API v2 documentation.
Trust Center: Learn about end-to-end encryption, create and manage your secret key, and see security details.
Create secret key: Click on Create secret key, enter it and confirm.
Make sure to keep your code in mind, we cannot recover it for you.
Help: Access support resources.
3.1 Start a Conversation
To start a conversation with an AI, simply click on its name.
Type your question in the chat box and receive an answer instantly.
If you want to attach a file, click on attach files, choose the files you want to add and confirm.
You can start to chat!
This option is only available for UnderstandAI.
Thinking Mode allows the AI to take extra time to reason through complex questions before responding. When enabled, the model performs deeper analysis, considers more possibilities, and produces more accurate, structured, and reliable answers—especially for multi-step or technical tasks.
Use this feature when your assistant needs to:
Solve complex problems
Analyze data or documents
Generate detailed explanations
Make structured decisions
This option is only available for Perplexity, GPT-4.1, and GPT-5.
The Search the Web feature allows your AI assistant to retrieve real-time information from the internet. When enabled, the model can look up recent facts, news, statistics, product details, or any data that may not be included in its training or knowledge base.
Use this feature when your assistant needs to:
Access up-to-date information
Check facts, prices, or recent events
Retrieve details from public websites
Enrich answers with the latest available data
see:
3.2 Compare AI Responses
To compare answers from two different AI models (e.g., Perplexity vs GPT-5):
Click the small “vs” icon that appears left next to the LLM.
Enter your question once.
Receive answers from both models.
3.3 Multi conversation
Each assistant has its own memory and can handle multiple conversations in parallel.
Conversations are topic-specific, allowing the assistant to maintain accurate and contextual history.
You can switch between conversations at any time without losing track of previous exchanges.
3.4 Chained Prompts
The Chained Prompts feature allows your AI assistant to execute a series of prompts in sequence, producing a complete, structured output from start to finish. This is useful for creating long documents, structured reports, or step-by-step workflows.
Use the Welcome Bot (Optional)
You can ask the Welcome Bot to generate a chained prompt template for your subject. Example:
“Write a chained prompt for an employment contract in France.”
Prepare your prompts in a .txt File
Write each prompt in your workflow using triple quotes (
""") for clarity.Separate each prompt with a comma.
Example for an employment document:
"""Summarize a standard French employment contract""",
"""List 5 key employee rights in the contract""",
"""Create a CSV of main clauses and their purpose"""Upload Your Chained Prompts File
In the chat section, choose the LLM you wish to use and select the chained prompts button.
Upload your
.txtfile containing the chained prompts.
Execute your Chained Prompts
The AI will execute your prompts sequentially, merging outputs in order to produce a complete, coherent document or workflow.
Once the prompts completed, download the file
You can choose either word or excel to download your document.
Last updated