Uploading & Working with Files in ThreoAI

Overview

ThreoAI lets you attach files directly to your conversations. Once uploaded, the AI can read, analyze, summarize, and answer questions about the content. This works with documents, images, spreadsheets, and more - and it is available in all chat contexts, including the main home chat, Project workspaces, and Custom GPT or AI Agent conversations.

File uploads allow you to go beyond text-based prompts and bring your actual documents, data, and images into the conversation. Whether you need to summarize a lengthy report, extract data from a spreadsheet, compare two contracts, or describe what an image contains, ThreoAI can process the uploaded content and respond based on what it finds.

---

Uploading a File

To attach a file to your current conversation:

1. Open a chat or start a new one. File uploads work in any chat context - a general conversation, a Project workspace, or a Custom GPT / AI Agent chat.
2. Click the + button in the chat input bar (located on the left side of the input field).
3. Select Attach files from the dropdown menu that appears.
4. A file picker dialog opens. Choose one or more files from your device. You can select multiple files at once if you need to upload several documents together.
5. The selected files appear as attachment thumbnails above the input bar, showing the file name and a preview icon.
6. Type your message in the input bar to tell the AI what you want it to do with the files. For example, “Summarize this document” or “Extract all line items and their costs from this invoice.”
7. Press Enter to send both your message and the attached files together.

The AI will process the uploaded files as part of your message and generate a response based on the file content and your instructions.

Supported File Types

ThreoAI supports a wide range of common file formats for upload and analysis:

Category	Supported Formats
Documents	PDF, DOCX, TXT, PPTX
Images	PNG, JPG, JPEG, GIF, WEBP
Spreadsheets	XLSX, CSV

Each format is handled differently by the AI:

• PDF files are read and processed as text. The AI can summarize content, extract specific sections, answer questions about the document, and compare multiple PDFs.
• DOCX files (Microsoft Word) are processed similarly to PDFs. The AI reads the text content and can analyze, summarize, or reformat it.
• TXT files (plain text) are read directly. These are useful for providing raw text data, code files, log files, or other unformatted content.
• PPTX files (PowerPoint presentations) are processed for their text content across slides. The AI can summarize the presentation, extract key points, or describe the content of individual slides.
• Images (PNG, JPG, JPEG, GIF, WEBP) are analyzed visually. The AI can describe what an image shows, extract text from screenshots or photos of documents (OCR), identify objects or diagrams, and answer questions about visual content.
• XLSX files (Excel spreadsheets) are read as structured data. The AI can analyze trends, summarize data, identify patterns, calculate totals, and answer questions about the numbers and categories in the spreadsheet.
• CSV files (comma-separated values) are processed as tabular data, similar to XLSX files. They are commonly used for data exports and are well-suited for analysis tasks.

File size limits are set by your organization’s configuration. If a file fails to upload, try splitting it into smaller parts or contact your administrator to check the configured limits.

---

Previewing Documents in Chat

After uploading a file and receiving the AI’s response, you can still review the original file directly within ThreoAI:

• Click the attachment thumbnail or file name in the chat to open a preview of the original file.
• PDF files render inline in the chat window, allowing you to scroll through pages without leaving ThreoAI. This is useful for verifying that the AI’s summary or analysis matches the source document.
• Images are displayed inline in the conversation at a viewable size. Click an image to see it in a larger view.
• Other file types may display a file icon with the option to download the original for review in the appropriate application on your device.

---

Downloading Files

Files shared in a conversation - either files you uploaded or files generated by the AI (such as code files, formatted outputs, or processed data) - can be downloaded directly from the chat:

1. Locate the file attachment or AI-generated file in the conversation thread.
2. Click the download icon or the file name to initiate the download.
3. The file saves to your device’s default downloads folder.

This is useful when the AI produces output you want to keep, such as a reformatted document, extracted data, or generated code.

---

Tips for Working with Files

Getting the best results from file analysis depends on how you frame your request to the AI. Here are some practical tips:

Be Specific in Your Prompt

Instead of a generic request like “analyze this file,” provide specific instructions about what you want the AI to focus on. Specific prompts produce more targeted and useful responses.

• Instead of: “Analyze this file.”
• Try: “Summarize the key findings in this quarterly report, focusing on revenue trends and operating expenses.”
• Instead of: “Look at this spreadsheet.”
• Try: “Calculate the total spending by category and identify which category had the largest month-over-month increase.”
• Instead of: “What does this say?”
• Try: “Extract all action items from these meeting notes and list them with their assigned owners and due dates.”

Working with Multiple Files

You can attach more than one file to a single message. This is particularly useful when you need to:

• Compare documents - Upload two versions of a contract and ask the AI to identify the differences.
• Cross-reference content - Upload a policy document and an audit report, then ask the AI to check whether the audit findings align with the policy requirements.
• Consolidate information - Upload several related documents and ask the AI to create a single summary that draws from all of them.

Working with Large Documents

For very long documents, the AI may not be able to process the entire file in full detail in a single pass. To get the best results with large documents:

• Ask the AI to focus on specific sections: “What does section 3 say about refund policies?”
• Request a high-level summary first, then drill into specific areas with follow-up questions.
• If a document is extremely large, consider splitting it into smaller files and uploading them one at a time.

Working with Images

ThreoAI can analyze images in several ways:

• Describe visual content - “What does this screenshot show?” or “Describe the layout of this diagram.”
• Extract text (OCR) - “Transcribe the text in this image” or “What does the error message in this screenshot say?”
• Analyze charts and graphs - “What trends are shown in this bar chart?” or “Summarize the data in this pie chart.”
• Identify objects - “What products are shown in this photo?” or “What type of device is pictured here?”

Working with Spreadsheets

When uploading XLSX or CSV files, you can ask the AI to perform a variety of analysis tasks:

• Summarize the data and identify key trends
• Calculate totals, averages, or other aggregate values
• Find outliers or anomalies in the data
• Compare data across different columns, rows, or time periods
• Generate insights or recommendations based on the numbers

For best results, make sure your spreadsheet has clear column headers so the AI can understand the structure of the data.

---

Files in Custom GPTs

When building a Custom GPT, you have the option to upload Training Documents (PDFs, DOCXs, TXTs) that become part of the GPT’s persistent knowledge base. This is different from attaching a file to a single conversation:

• Conversation file attachments are one-time uploads. The AI processes them for the current conversation only.
• Custom GPT training documents are uploaded during GPT creation and become a permanent part of the GPT’s knowledge. Every conversation with that GPT can draw from the training documents automatically, without the user needing to re-upload anything.

This makes Custom GPT training documents ideal for building assistants that answer questions from a specific set of documentation - such as company policies, product manuals, SOPs, or knowledge base articles.

See Creating a Custom GPT for full instructions on uploading training documents and configuring a GPT’s knowledge base.

---

Troubleshooting File Uploads

If you encounter issues when uploading or working with files, here are some common problems and solutions:

File fails to upload.

• Check that the file type is supported (PDF, DOCX, TXT, PPTX, PNG, JPG, JPEG, GIF, WEBP, XLSX, CSV).
• Check that the file does not exceed your organization’s size limit. If you are unsure of the limit, contact your administrator.
• Try splitting a large file into smaller parts and uploading them separately.

The AI does not seem to read the file content correctly.

• Make sure the file is not password-protected or encrypted. The AI cannot read protected files.
• For scanned PDFs (images of text rather than actual text), the AI may have difficulty extracting content. Try using a higher-quality scan or converting the document to a text-based PDF.
• Rephrase your prompt to be more specific about what section or content you want the AI to focus on.

The AI’s summary seems incomplete for a large document.

• Large documents may exceed the model’s processing capacity in a single pass. Try asking about specific sections or pages rather than requesting a summary of the entire document at once.
• Break your analysis into multiple follow-up questions to cover different parts of the document.

---

Related

• Starting a Chat - how to initiate conversations and use the chat input bar
• Creating a Custom GPT - upload training documents as a GPT’s persistent knowledge base
• Navigating the Interface - learn about the + button and chat input bar features
• Projects - organize file-heavy conversations into dedicated workspaces