Direct answer
You change the ChatGPT model by opening a chat, clicking the model name shown at the top of the conversation (next to the chat title), and selecting a different model from the dropdown or selector sheet. On mobile, tap the model name in the header to open the same selector. The list you see is not universal: it depends on your plan, your workspace settings, and which models OpenAI currently offers [1].
How Do You Change the ChatGPT Model on Web, Desktop, and Mobile?
The mechanism is the same everywhere — you are not editing a setting buried in preferences, you are changing a per-conversation control. On the web and in the desktop app, the current model is displayed at the top of the chat area; clicking it opens a menu of the models your account can access, and the next reply you generate uses whichever one you pick [2]. Nothing is installed or restarted, and the change applies to new messages rather than rewriting earlier ones.
On iOS and Android, the model name sits in the header of the conversation. Tapping it opens a bottom selector sheet listing the available models, and choosing one closes the sheet and returns you to the same thread [2]. If you are on a small screen and cannot find it, check that you are inside an active chat rather than the chat history list — the selector only appears once a conversation is open.
It helps to know what switching does not do. Your earlier messages stay in the thread, but they were produced by the previous model, so a long conversation can end up mixing outputs from two different systems. For anything you will submit for assessment, start a fresh chat after switching so the whole draft comes from one model and is easier to reason about later. The desktop app mirrors the web picker closely, so instructions you learn on one carry over to the other [2].
One practical caution: model names change frequently, and a model you used last semester may now sit under a "legacy" or "older models" grouping, or may have been retired entirely. If a tutorial you are following names a model you cannot see, treat that as a version difference rather than a mistake on your part, and pick the closest current equivalent [1].
Why Is the Model Switcher Missing or Greyed Out in ChatGPT?
Most "the option isn't there" cases are entitlement, not malfunction. Free and paid tiers expose different model lists, so a free account may show only a small subset while a Plus, Pro, Business, or Edu account shows more [3]. If a classmate can switch to a model you cannot see, comparing plan tiers usually explains the gap before you assume a bug.
Workspace controls are the second common cause. In Business, Enterprise, and Edu workspaces, administrators can enable or disable specific models for everyone in the organisation, and an individual user cannot override that from the picker [3]. If you are signed into a university-managed workspace, the model list you see may be deliberately narrower than your personal account's, and the fix is a request to your admin rather than a settings change on your side.
Account status and availability matter too. Regional rollout differences, an expired or failed subscription payment, and models that have been retired from the product can all remove options from the menu [3]. A greyed-out entry usually means the model exists in the interface but your account is not currently entitled to use it, which is different from an entry that has disappeared altogether.
Before you troubleshoot further, do the cheap checks in order: confirm you are signed into the account you think you are, confirm you are inside an open chat, refresh or update the app, and then compare your plan against the model you are trying to reach [3]. If all four line up and the option is still absent, the restriction is almost certainly on the account or workspace side rather than in your browser.
Does Switching ChatGPT Models Change How Turnitin's AI Detector Scores Your Writing?
No — Turnitin's AI writing detection does not identify which model generated a passage, so switching models does not change the scoring rules applied to your submission. The detector analyses the text you actually submit, segmenting it and reporting the proportion it classifies as likely AI-generated [4]. Model choice can influence how a draft reads, and therefore how it is classified, but there is no "GPT-4o penalty" or "GPT-5 exemption" in the report.
What the report gives you is a percentage plus highlighted spans, not an attribution. Turnitin has been explicit that its indicator is a signal to be interpreted alongside context and conversation with the author, not standalone proof of misconduct [4]. That framing matters for students: a high score is a prompt to review the flagged passages and be ready to discuss your process, while a low score is not a guarantee that an instructor will accept the work without question.
There is also a display detail worth understanding before you panic at a report. When the AI signal falls below Turnitin's confidence threshold, the score is shown as *% rather than an exact figure — a low-confidence indication rather than a precise measurement [4]. Students frequently misread that asterisk as an error or a hidden high score, when it generally indicates the opposite.
The practical takeaway is that model switching is a drafting decision, not a compliance strategy. If you rewrote AI-assisted prose in your own words, kept your notes and drafts, and can explain your sources, you are in a far stronger position than someone gaming model names. Checking your own draft against a Turnitin report before the real deadline turns an anxious guess into information you can act on [4].
Switching models is the easy part; knowing what the resulting draft will look like in a real Turnitin report is where most students get caught out. turnitin0 lets you upload your .docx, .pdf, or .txt file and receive the same AI detection and similarity reports your professor sees — typically in under 15 minutes, with no subscription and no addition to Turnitin's student paper database.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Where exactly is the model selector in ChatGPT?
It sits at the top of an open conversation, next to the chat title on web and desktop, and in the header on mobile; tapping or clicking the current model name opens the list of models your account can use [1].
Why can I only see one or two models?
Model availability follows your plan tier and any workspace restrictions set by an administrator, so a narrower list usually reflects entitlement rather than a technical fault [3].
Does changing the model mid-conversation affect earlier messages?
No. Previous replies remain as they were written, but new replies use the newly selected model, which is why starting a fresh chat is cleaner for anything you plan to submit [2].
Will a different model make my text score lower in Turnitin?
Not directly. Turnitin's AI writing detection reports on the submitted text itself and does not attribute passages to a specific model, so the score reflects how the writing reads rather than which tool produced it [4].
What does the *% score in a Turnitin AI report mean?
It means the AI signal fell below Turnitin's confidence threshold, so an exact percentage is not displayed; it is a low-confidence indication rather than a hidden high score [4].