Key takeaways
- A ChatGPT project groups related chats, files and custom instructions so context carries across conversations.
- Use projects for ongoing work, such as a course, client or app, and plain chats for one-off questions.
- Shared projects let collaborators work with the same files and instructions on supported plans.
- ChatGPT Atlas is OpenAI's web browser with ChatGPT built in; plugins were replaced by GPTs and apps.
- ChatGPT was trained by pretraining on large text datasets, then fine-tuning with human feedback.
A ChatGPT project is a workspace that groups related chats, uploaded files and custom instructions, so ChatGPT keeps the same context across many conversations about one piece of work, such as a semester course, a client account or an app you are building. Projects make long-running work far more organised than scattered chats. I build AI products and use AI workspaces daily, so this page covers projects, using and sharing them, collaborating, ChatGPT Atlas, how ChatGPT is trained, plugins, and fixing ChatGPT when it stops responding.
What is a ChatGPT project?
A ChatGPT project is a folder-like workspace in ChatGPT that holds chats, files and instructions for one topic. Chats inside a project can use the project's files and instructions, and ChatGPT can reference earlier project conversations, so you do not repeat background every time.
Also asked as: chatgpt project vs chat
Typical uses: studying a subject with lecture notes uploaded, managing a client with their brief and documents, planning an event, writing a long report, or developing software with specifications. Projects are similar in purpose to Claude's projects and Gemini's Gems with files: reusable context for focused, ongoing work. For students, one project per subject, with the syllabus and notes uploaded, turns ChatGPT into a consistent study assistant that remembers the course context across weeks.
How do I use projects in ChatGPT?
Create a new project from the sidebar, name it, add files such as notes or documents, write project instructions describing the goal, audience and preferred style, then start chats inside it or move existing chats into it.
Also asked as: how to use projects in chatgpt
Tips:
- Keep each project focused on one goal.
- Write instructions like a brief: role, context, rules and output format.
- Upload only relevant, up-to-date files and remove outdated ones.
- Name chats clearly so you can find them later.
- Ask ChatGPT to summarise progress periodically within the project.
Projects work best when instructions and files are curated, not dumped.
What is the difference between a ChatGPT project and a chat?
A chat is a single conversation; once you start a new chat, earlier context is not automatically included, apart from any memory features. A project groups many chats with shared files and instructions, so context persists across conversations on the same topic.
Use regular chats for quick, unrelated questions. Use projects when you will return to the same work repeatedly, need consistent instructions, or rely on reference documents. Projects also keep your sidebar organised. For one-time tasks, a project adds unnecessary setup; for semester-long courses or ongoing client work, it saves significant time.
Can a ChatGPT project be shared?
Yes, on supported plans, OpenAI offers shared projects, letting you invite collaborators who can use the same files, instructions and project chats. Availability and permissions depend on your plan and workspace settings, so check OpenAI's current help article.
Before sharing, review what is inside: files, chat history and instructions may contain information others should not see. For team accounts, administrators can control sharing. If shared projects are not available on your plan, you can share individual chats through links, or create a custom GPT with the same instructions and files for others to use.
How do I share a project in ChatGPT?
Open the project, use its share or invite option, and add collaborators by email or link according to your plan's options, choosing their permissions. Collaborators then see the project's files, instructions and shared chats.
Also asked as: how to share a project in chatgpt
If you only need to share results, share a specific chat link instead, which gives read access to that conversation. Remember that shared content may include uploaded files and details from conversations. For sensitive work, use team or enterprise plans with admin controls and data protections rather than personal accounts.
How do I collaborate using a ChatGPT project?
Agree on the project's purpose, keep shared instructions and files current, use clear chat names for different workstreams, and record decisions in a summary document inside the project so everyone works from the same context.
Also asked as: how to use chatgpt project and collaborate
Collaboration tips:
- Assign an owner who maintains instructions and files.
- Keep one "status" chat or document summarising decisions.
- Upload final versions only, with dates in file names.
- Review AI-generated content before sharing it outside the team.
- Avoid uploading confidential data unless your plan's data controls allow it.
Projects support teamwork, but human review and clear ownership keep the work accurate.
What is ChatGPT Atlas?
ChatGPT Atlas is OpenAI's web browser with ChatGPT built in, launched in 2025, starting with macOS. It lets you ask ChatGPT about the page you are viewing, use optional browser memories for context, and run agent mode to complete tasks on websites.
Atlas integrates assistance directly into browsing: summarising articles, comparing products across tabs, drafting replies and filling forms with agent mode under your supervision. As with any AI browser, review privacy settings, be careful with agent access to logged-in accounts, and watch for prompt injection on untrusted websites. Check OpenAI's site for current platform availability. Atlas is part of a wider shift towards AI browsers, alongside Copilot in Edge and Gemini in Chrome, where assistants work directly on the pages you visit.
How is ChatGPT trained?
ChatGPT's models are trained in stages: pretraining on large amounts of text to learn language and knowledge by predicting the next token, then supervised fine-tuning on example conversations written by people, then reinforcement learning from human feedback, where people rank responses and the model learns to prefer helpful, safe answers.
Newer models add training for reasoning, tool use, safety and multimodal inputs, and draw on licensed, public and human-created data. Training happens before release, so models have a knowledge cut-off; web search and file uploads supply newer or private information at use time. OpenAI's system cards describe training and safety work for each model.
What were ChatGPT plugins?
ChatGPT plugins, introduced in 2023, let ChatGPT call external services, such as travel booking, shopping and data tools, through APIs. OpenAI phased them out in 2024 in favour of custom GPTs with actions, and later apps and connectors integrated into ChatGPT.
Also asked as: what are chatgpt plugins
The idea lives on: GPTs can call APIs through actions, and connectors link ChatGPT to services such as cloud storage and business apps. For developers, building on current options means defining actions with OpenAPI schemas or using newer app frameworks OpenAI provides. Old plugin tutorials are outdated, so follow current documentation.
Why is ChatGPT not responding?
ChatGPT may not respond because of service outages, high demand, usage limits, long conversations, large uploads, browser extensions, network problems, or an outdated app. Check OpenAI's status page first to rule out incidents.
Also asked as: why chatgpt is not responding
Fixes:
- Refresh the page or restart the app.
- Start a new chat if the conversation is very long.
- Disable extensions or try a private window.
- Clear the browser cache or update the app.
- Check your internet connection or switch networks.
- Wait if you have hit a usage limit.
If one model fails, switching to another model sometimes works while issues are resolved.
Sources
GenAI engineer and AI product builder with 2+ years shipping production-grade AI systems: RAG pipelines, fine-tuned LLMs, hybrid retrieval and multi-modal apps across vision, speech and OCR, architected end to end from ingestion to deployment. Leads engineering for SCULT INDIA's 14-member team, founded the 500+ member TechVerse Enclave community and has mentored 200+ students. Three hackathon first prizes: Changethon 2025 (IIT Roorkee), Product Genesis at Vividhotsava 2025 (CSJMU) and BYTEBATTLE (MeetKats).
Open to GenAI roles, hackathon judging, mentorship sessions and guest talks at colleges. For campus sessions the inviting institution covers the visit end to end: travel (a First AC train ticket for journeys up to five hours, a flight for anything longer), accommodation, meals, and pickup and drop; the honorarium and workshop fees are negotiable on WhatsApp or on the call, and requests are welcome from institutions that can cover the visit this way. Email pranjulrathour41@gmail.com.
