Choosing the right AI model for your task
19 June 2026
AI Tool
ChatGPT
Learn the differences between a standard and reasoning AI model, and when to use each of them.
What are Standard and Reasoning models?
Most people use ChatGPT for writing, summarising, or answering questions. For many of these tasks, a standard model (known as 'Instant' in ChatGPT) is fast and effective.
Reasoning models spend more time working through a problem before producing a response. In ChatGPT, you can adjust the level of thinking to "High" and above.
How do I decide which to use?
Here's a simple way to decide between a standard and reasoning model.
Or you can think of it this way:
Use a Standard Model for: | Use a Reasoning Model for: |
Rewriting text | Comparing options |
Drafting emails | Making recommendations |
Summarising documents | Building plans |
Creating meeting notes | Evaluating trade-offs |
Here are some examples of everyday scenarios
How much does it cost?
Different reasoning models cost different amount of credits. For ChatGPT you can use the table below as guide:
Model | Use it for: | Credits | Typical Wait |
|---|---|---|---|
Instant | Everyday tasks: emails, summaries, rewriting | Unlimited | Almost immediate |
Medium | Reports, planning, coding, and tougher questions | ~10 credits | <1min |
High | Complex analysis and multi-step work | ~10 credits | 1-2min |
Extra High | High-accuracy tasks where mistakes are costly | ~10 credits | 2-5min |
Pro | Mission-critical and expert-level problems | ~50 credits | >5min |
You can check your amount of credits left by going to 'SG WOG ChatGPT' - 'Settings' - 'Usage'
Now it's your turn to try
Choose one recurring task from your work
Ask yourself: would a standard or reasoning model be more appropriate?Try both models
Run the same task through both a standard and a reasoning model, and compare the results.