Use AI more efficiently

Why this matters

Every prompt may use part of your monthly AI credit allowance. Using credits wisely helps you get faster responses, better results, and more value from your allocation.

1) Choose the right model

Not every task requires the most capable reasoning model. Simple tasks such as summarising emails often perform well using smaller, faster models.

Using the appropriate model can significantly reduce costs while maintaining quality.

2) Use Projects to keep context

One of the biggest sources of wasted tokens is repeatedly providing the same background information.

Instead of copying your instructions into every new conversation, use Projects (where available) to store writing style, background information and standard instructions.

A Project in ChatGPT

3) Retrieve only what you need

Only provide documents that are relevant to the task.

When using connectors such as SharePoint, ask AI to search for specific files rather than your entire Sharepoint. This keeps responses focused and avoids processing unnecessary information.

4) Start focused

The clearer your request, the fewer follow-up questions the AI needs to ask. A useful structure is to include three pieces of information up front:

5) Keep the conversation going

Continue related work in the same chat so the AI can reuse existing context.

6) Save and reuse prompts

Save prompts you use regularly for tasks such as:

• Summarising reports

• Submission writing

• Email drafting

Reusable prompts improve consistency and reduce repetitive prompting.

Key takeaways: