Starting a Chat in ThreoAI
How to start a conversation in ThreoAI - send messages, switch AI models, set a default model, enable web search, chat with Custom GPTs and AI Agents, and manage your conversation history.
Overview
Section titled “Overview”Chatting in ThreoAI works like any AI chat tool - type a message, press Enter, and the AI responds. What makes ThreoAI different is that the AI models available to you are curated and managed by your organization, ensuring they meet your company’s security and compliance requirements.
This guide covers everything you need to know to start conversations, choose models, enable web search, and manage your chat history.

Starting a New Chat
Section titled “Starting a New Chat”There are several ways to start a fresh conversation in ThreoAI:
- Click the New chat button in the top-left area of the sidebar. This creates a new conversation in your general chat history.
- Use the keyboard shortcut Ctrl + Shift + O (or Cmd + Shift + O on Mac) to start a new chat from anywhere in the interface.
- Navigate directly to https://threo.synthreo.ai/#/home in your browser to return to the home screen and start fresh.
Once you are on the home screen or in a new chat:
- Type your message in the input bar at the bottom of the screen.
- Press Enter to send your message. The AI will begin generating a response immediately.
Use Shift + Enter to insert a line break in your message without sending it. This is helpful when you want to write a multi-line prompt, include structured instructions, or format your message before sending.
Tip: The quality of the AI’s response depends heavily on how you phrase your prompt. Be as specific and clear as possible about what you want. Instead of a vague prompt like “help me with this,” try something more targeted like “summarize the three main arguments in this document and list the supporting evidence for each.”
Choosing an AI Model
Section titled “Choosing an AI Model”The model selector in the top-right of the screen shows the currently active AI model for your conversation (for example, GPT 5.4, Claude Sonnet 4.6, or another model). Click it to open a dropdown menu showing all available models.

Select any model from the dropdown to switch to it for the current conversation. The model change takes effect immediately - your next message will be processed by the newly selected model.
The models available to you are configured by your organization’s administrator in Tenant Management. Different models have different strengths - some are better at creative writing, others at technical analysis, and others at following complex instructions. If you need access to a model that isn’t listed, contact your admin to have it enabled.
Setting a Default Model
Section titled “Setting a Default Model”If you find yourself switching to the same model every time you start a new chat, you can save time by setting a default model in your Profile Settings. To do this:
- Click your profile avatar in the top-right corner and select Profile.
- On the General tab, find the Default Model dropdown.
- Select the model you want to use by default for all new conversations.
- Click Save.
From this point on, every new chat will start with your chosen default model already selected. You can still switch models manually during any conversation - the default only affects which model is selected when you first start a new chat.
If you leave the Default Model field empty, ThreoAI will use your organization’s default model instead.
Enabling Web Search
Section titled “Enabling Web Search”By default, ThreoAI answers from the AI model’s built-in training knowledge, which has a knowledge cutoff date. When you need information that may have changed since the model’s training data was collected - such as recent news, current pricing, the latest version of a software product, or any time-sensitive information - you can enable Web Search.

To toggle Web Search on or off:
- Click the + button in the chat input bar (this button is available in any chat context - home chat, Projects, Custom GPTs, and AI Agents).
- Select Web Search from the dropdown menu to enable it.
- Send your message. ThreoAI will search the web for relevant, up-to-date information and include the results in its response alongside its built-in knowledge.
To turn Web Search off again, repeat the same steps - click the + button and select Web Search to toggle it back to disabled.
When to use Web Search:
- When you need current or real-time information (news, stock prices, weather, recent events)
- When you are asking about something that may have changed after the model’s training cutoff date
- When you want the AI to verify its answer against current online sources
When Web Search is not needed:
- For general knowledge questions that do not depend on recency
- When working with files you have uploaded (the AI analyzes your files directly)
- When chatting with a Custom GPT that has its own knowledge base of uploaded training documents
Chatting with a Custom GPT or AI Agent
Section titled “Chatting with a Custom GPT or AI Agent”Custom GPTs and AI Agents work the same as a standard chat, but with a custom system prompt, persona, and optionally a knowledge base pre-loaded. This means the AI is already configured to behave in a specific way and may have access to specific documents or data that a general chat does not.
Finding and Opening a Custom GPT
Section titled “Finding and Opening a Custom GPT”- From the sidebar: Look in the My GPTs section of the left sidebar. Click any GPT listed there to open a chat with it.
- From My GPTs page: Click My GPTs at the bottom of the My GPTs sidebar section (or navigate to
#/my-gpts) to see all your GPTs in a card layout. Click any card to open the GPT. - From the Marketplace: Click Explore Marketplace in the sidebar (or navigate to
#/marketplace). Use the Custom GPTs filter tab to browse GPTs shared by your organization. Click any GPT card to open its detail panel, then click Start Chat.
Finding and Opening an AI Agent
Section titled “Finding and Opening an AI Agent”- From the sidebar: Look in the AI Agents section of the left sidebar. Click any agent listed there to start a conversation immediately.
- From the Agents directory: Click Explore Agents at the bottom of the AI Agents sidebar section to open the full agents directory at
#/explore-agents. Browse or search for agents, then click Use Agent to start a conversation. - From the Marketplace: Click Explore Marketplace in the sidebar. Use the AI Agents filter tab to browse agents. Click any agent card to open its detail panel, then click Start Chat.
Using Conversation Starters
Section titled “Using Conversation Starters”If the GPT or Agent has conversation starters configured by its creator, they appear as suggested prompts when you first open it. These are pre-written prompts designed to demonstrate what the GPT or Agent does best.
- Click any conversation starter to send it immediately as your first message.
- Alternatively, ignore the starters and type your own message in the input bar.
See Using a Custom GPT and Agents & Marketplace for more detail on getting the best results from GPTs and Agents.
Attaching Files to a Chat
Section titled “Attaching Files to a Chat”You can attach files to any conversation for the AI to analyze. This works in all chat contexts - the main home chat, Project workspaces, and Custom GPT or AI Agent conversations.
- Click the + button in the chat input bar.
- Select Attach files from the dropdown menu.
- Choose one or more files from your device.
- The selected files appear as thumbnails above the input bar.
- Type your message describing what you want the AI to do with the files, then press Enter to send.
Supported file types include PDF, DOCX, TXT, PPTX, images (PNG, JPG, JPEG, GIF, WEBP), XLSX, and CSV. For detailed guidance on working with files, including tips for getting the best results, see Uploading & Working with Files.
Managing Your Chat History
Section titled “Managing Your Chat History”All conversations are saved automatically as soon as you send your first message. You do not need to manually save anything - your entire conversation history is preserved and accessible from the left sidebar.
Finding Past Conversations
Section titled “Finding Past Conversations”Past conversations appear in the Previous section of the left sidebar, sorted by most recent. Scroll through the list to find a past chat, or use the Search feature (click the search icon at the top of the sidebar, or press Ctrl + K / Cmd + K) to search for a conversation by keyword.
Renaming a Conversation
Section titled “Renaming a Conversation”By default, ThreoAI assigns a name to each conversation based on its content. To rename a conversation to something more descriptive:
- Hover over the conversation in the sidebar to reveal the … (three-dot) menu.
- Click … and select Rename.
- Type a new name and confirm.
Deleting a Conversation
Section titled “Deleting a Conversation”To remove a conversation from your history:
- Hover over the conversation in the sidebar to reveal the … (three-dot) menu.
- Click … and select Delete.
- Confirm the deletion when prompted.
Note: Deleting a conversation is permanent and cannot be undone.
Organizing with Projects
Section titled “Organizing with Projects”If your chat history is getting long and hard to navigate, consider using Projects to organize related conversations into separate workspaces. Conversations inside a project do not appear in the Previous section of the sidebar, which keeps your general history clean and focused.
Tips for Effective Conversations
Section titled “Tips for Effective Conversations”Here are some best practices to help you get the most useful and accurate responses from ThreoAI:
- Be specific. The more detail you include in your prompt, the better the AI can tailor its response. Instead of “write an email,” try “write a professional email to a client explaining that their project deadline has been extended by two weeks.”
- Provide context. If the AI needs background information to answer your question well, include it in your message. For example, “I’m an IT support specialist. A user is reporting that their Outlook keeps crashing after the latest Windows update. What troubleshooting steps should I suggest?”
- Break complex tasks into steps. For multi-part requests, ask the AI to handle one part at a time rather than everything in a single prompt.
- Use follow-up messages. If the AI’s first response is not quite right, send a follow-up message to refine it. For example, “Make that more concise” or “Focus on the cost comparison section.”
- Set your personal context. Configure your Profile Settings with your role, communication preferences, and any other relevant context. This information is included with your conversations, so the AI can give more relevant and personalized responses without you having to repeat it every time.
Related
Section titled “Related”- Navigating the Interface - full interface reference
- Uploading & Working with Files - attach documents and files to your conversations
- Comparing AI Models - run the same prompt against two models side by side
- Projects - group conversations into workspaces
- Using a Custom GPT - how to find and use Custom GPTs
- Profile Settings - set your default model and personalize your experience