Why plan and model choices matter at work
In Level 1 you saw that ChatGPT can sound confident and still be wrong, and that what happens to your data depends on your settings. Two choices control a lot of that before you type anything:
- The plan decides your usage limits, which tools you get, whether there is a shared team workspace, and whether your chats can be used to train OpenAI's models by default.
- The model or thinking level you pick for each chat decides how fast the answer comes back and how carefully ChatGPT works through the problem.
Think of the plan as the vehicle your company leases and the model picker as the gear you choose for each journey.
Task 1: Build a plan comparison table
OpenAI changes plans, prices and limits often. Always work from the live pricing page at openai.com/chatgpt/pricing, not from a blog post or from memory.
Do this:
- Open the pricing page and switch between the Personal and Business views.
- Make a table with one row per plan: Free, Go, Plus, Pro, Business, Enterprise.
- Add these columns: Price, Usage limits, Tools included, Workspace features and Training on your data by default?
- Fill in each cell from the page. Write "not stated" if the page does not say.
- Put the date at the top of the table. It will go out of date.
As of October 2026, a filled-in table looks roughly like this:
| Plan | Price (US) | What it changes for work | Training by default |
|---|---|---|---|
| Free | $0 | Basic access with tight limits; may show ads | On (you can switch it off) |
| Go | $8/month | Higher limits than Free for uploads, images and data analysis; projects and custom GPTs; may show ads | On (you can switch it off) |
| Plus | $20/month | Higher limits, more thinking levels, no ads | On (you can switch it off) |
| Pro | Several tiers, from $100/month | Highest personal limits and the top thinking levels | On (you can switch it off) |
| Business | $25/user/month monthly, $20 billed annually; minimum 2 seats | Shared workspace, admin console, SSO, connectors controlled by admins | Off |
| Enterprise | Custom, through OpenAI sales | Business features plus enterprise-grade controls and support | Off |
Note the training column. On personal plans (Free, Go, Plus, Pro) chats can be used for training unless you turn off Settings > Data controls > Improve the model for everyone. On Business and Enterprise, OpenAI does not train on your inputs and outputs by default. For a team handling client or staff data, this column often settles the decision.
Task 2: Choose a plan for a team scenario
Scenario: A five-person HR team wants to draft job adverts, summarise policy documents stored in SharePoint and share prompts with each other. Some documents contain employee details.
Do this:
- List the team's needs: shared workspace, connection to SharePoint, no training on company data, admin control.
- Go through your table row by row and cross out any plan that fails a need.
- Free, Go, Plus and Pro are personal plans: no shared admin console, and training is on by default.
- Business includes an admin console and connectors such as SharePoint and Google Drive, and does not train on your data by default.
- Pick the cheapest plan left. Here that is Business. Enterprise is worth a sales call only if the company needs more controls or support than Business offers.
- Write a two-line justification: the need, and the plan feature that meets it.
- Check for flags. Before you recommend a plan, look for features marked beta, preview or not available in all regions. Ads, for example, are being rolled out on Free and Go only in selected regions. Write any such feature into your justification so nobody counts on something they may not get.
Also check your company's AI policy. Some employers allow only an approved company workspace.
Task 3: Pick a model for each task
The model picker sits at the top of the chat window. Its labels change often. Recently it has offered thinking levels such as Instant (fast everyday answers), Medium, High, and on higher plans Extra High and Pro for the hardest work. Free and Go users get a lighter default model with fewer options. Your options depend on your plan and on your workspace admin's settings.
Rule of thumb:
- Quick task (reword an email, write a subject line, give a short definition): use the fastest option.
- Complex task (weigh three vendor quotes against criteria, plan a project with dependencies): use a higher thinking level. It takes longer and usually reasons more carefully.
Do this:
- Open a new chat and select the fastest option.
- Ask: "Rewrite this sentence to sound more polite: 'Send the report by Friday.'" Note roughly how long it took.
- Open another chat and select a higher thinking level.
- Ask: "A supplier offers 5% off if we pay in 10 days instead of 30. Explain whether that is a good deal and what we should consider." Note the time.
- Run the supplier question with the fast option too, then compare the two answers.
- Record the results in a small log: task, option used, time taken, and quality (was it correct, complete and well reasoned?).
Check your result: You should usually find little or no quality gain from the slower option on the quick task, and a clearer, better-reasoned answer from it on the complex task. If not, write down what you saw. That is useful evidence for your team.
Key takeaways
- Plans and limits change often. Work from OpenAI's live pricing page and date your comparison table.
- Business and Enterprise do not train on your data by default. On personal plans it is on until you switch it off.
- Choose the cheapest plan that meets every team need, and write down any beta or region-limited feature.
- Use fast options for simple tasks and higher thinking levels for complex reasoning.
- Test, time and log your model choices instead of guessing.