AI Model Wiki

    Skip to main content
    10 min read
    Lesson 1 of 5

    Understanding ChatGPT Projects

    From the course ChatGPT Projects & Custom GPTs

    ChatGPT Projects let you organize conversations, files, and custom instructions into dedicated workspaces. Instead of losing context across scattered chats, Projects keep everything in one place.

    What Are Projects?

    A Project is a persistent workspace that includes:

    • Conversations — All related chats grouped together
    • Files — Upload documents, images, data that persist across conversations
    • Custom Instructions — Project-specific behavior and context
    • Shared Context — Every new chat in the project inherits its settings

    Creating a Project

    1. Click "New Project" in the sidebar
    2. Name it descriptively (e.g., "Q1 Marketing Campaign")
    3. Add custom instructions for the project
    4. Upload relevant files
    5. Start conversations within the project

    Project Instructions

    Set instructions that apply to all conversations in the project:

    • "You are a marketing strategist for a B2B SaaS company targeting enterprise clients."
    • "Always respond in British English. Use metric units."
    • "Reference the uploaded brand guidelines document for tone and style."

    Organizing by Use Case

    ProjectInstructionsFiles
    Blog WritingBrand voice guide, SEO requirementsStyle guide, keyword list
    Code ReviewsLanguage standards, architecture docsCoding conventions PDF
    Customer SupportProduct knowledge, tone guidelinesFAQ document, product specs
    ResearchAcademic style, citation formatResearch papers, data sets

    Files in Projects

    Upload files that persist across all conversations:

    • PDFs, documents, spreadsheets
    • Images and screenshots
    • Code files and repositories
    • Data files (CSV, JSON)
    • Up to 20 files per project with generous size limits

    Benefits Over Regular Chats

    • No re-explaining context in every conversation
    • Files available without re-uploading
    • Consistent AI behavior across all project chats
    • Easy to find and resume related work

    Named Example: HubSpot's Content Team

    HubSpot's blog team publicly documented using ChatGPT Projects to maintain a "Brand Voice Project" containing their style guide, audience personas, and past top-performing articles. Every new blog draft starts in the project — ensuring the AI consistently matches HubSpot's conversational, data-driven editorial style without editors having to re-paste guidelines into every session.

    Example Project Instruction Template:
    "You are a content strategist for [Company]. Always:
    1. Write in second person, direct and conversational
    2. Lead with a concrete statistic or question
    3. Reference our product [ProductName] where relevant
    4. Target reading level: 8th grade (Flesch-Kincaid)
    See attached brand_guide.pdf for full style reference."
    

    Common Pitfall: Projects with too many uploaded files (10+) slow down response generation because the AI must process more context. Keep project files to the 3-5 most essential documents and refresh them quarterly rather than accumulating everything over time.

    Key Takeaways

    • •Projects group conversations, files, and instructions into persistent workspaces
    • •Custom instructions apply to all conversations within a project
    • •Uploaded files persist across conversations without re-uploading
    • •Organize projects by use case for maximum productivity
    0 of 5 lessons marked complete

    Progress is kept in this browser only.