Introducing User Budgets in FlyerGPT
To keep FlyerGPT sustainable and available to everyone across campus, we've introduced individual usage budgets. Every time you log in, you'll now see a budget monitor graphic that shows exactly where you stand for the month when you click on it. No more guessing about how much you've used.
What Are the Limits?
- Students: $20/month
- Faculty & Staff: $25/month
These budgets reset monthly and are designed to comfortably cover typical day-to-day use of FlyerGPT when using OneChat and Agents.
Need a Higher Limit?
If your work requires more headroom, for example, heavy research, large document processing, or coursework with intensive AI use, you can request a budget increase for that month. Just submit an AI Support Request through TeamDynamix, and the AI Applications & Services team will review it. Submit an AI Support Request
Best Practices for Lowering Your Monthly Token Cost
A few simple habits can meaningfully stretch your budget:
- Choose the right model for the task. FlyerGPT gives you a model choice; save the most powerful option for work that needs it, heavy research, complex documents, coursework with intensive AI use, and use a lighter one for quick lookups and simple formatting.
- Start new conversations for new topics. FlyerGPT resends the full conversation history with every message; nothing gets summarized or forgotten. A conversation that's drifted from topic A to topic B to topic C is quietly re-paying for A and B every time you ask about C. Close out finished topics and start a fresh chat.
- Use skills instead of re-explaining yourself every time. If you find yourself re-pasting the same rubric, template, or instructions into OneChat over and over, ask the AI Applications & Services team whether a skill already exists for that workflow, or could. Skills load their full instructions only when they're relevant to your request, so you don't pay to resend the same setup in every conversation.
- Combine related questions into one message instead of several. Because history compounds, five short back-and-forth messages usually cost more than one message that asks all five things up front; you're paying for the accumulated thread either way. Front-load your context and questions together.
- Be specific and concise in your prompts. Clear, focused asks get usable answers with less back-and-forth clarification, fewer turns, and less history to accumulate.
- Ask for the length you actually need. Open-ended requests tend to produce longer, more elaborate responses than ones that specify "one paragraph" or "just the key points."
- Upload only what's relevant. Trim a long document to the section you need before attaching it to FlyerGPT; the whole file counts against that conversation's budget even if you only ask about one part.
- Only invoke agents when the task calls for them. Agents that connect to campus systems (Salesforce, Banner, Box, and others) do real work behind the scenes, which adds to what's processed alongside your message. Reach for them when you need that system access, not as a default for every question.
More to Come
Please subscribe to the blog to get future posts in this series, which will cover new features as well as more tips and tricks to help you get the most out of your AI tools on campus.