AI Tokens and Usage
Every interaction with your AI agent consumes tokens — the units that AI models use to process and generate text. UltraClaw handles all the complexity for you; all you need to know is how your credits work.What Are Tokens?
Tokens are the fundamental units AI models use to read and write text. A token is roughly ¾ of a word — so a typical message might use a few hundred tokens for both your input and the agent’s response. You don’t need to think about tokens directly. UltraClaw tracks everything and shows your usage in dollar terms on your dashboard.Included Credits
All plans include $15/mo in AI credits — Starter, Pro, and Dedicated alike. This covers hundreds of typical messages per month.
Top-Up Credits
Need more than your monthly allowance? Top up anytime:- 1 of credit
- Minimum purchase: $10
- Top-up credits never expire — they carry forward until you use them
Open Your Dashboard
Go to ultraclaw.ai/dashboard and navigate to Billing.
What Affects Token Usage
Not all messages cost the same. Here’s what influences consumption:| Factor | Impact |
|---|---|
| Message length | Longer messages (yours and the agent’s) use more tokens |
| Model choice | Some models cost more per token than others |
| Thinking level | Higher thinking levels use more tokens for deeper reasoning |
| Context window | Conversations with more history loaded consume more per turn |
| Integrations | Tool calls (calendar reads, email sends, etc.) add tokens |
Monitor Your Usage
Your dashboard shows:- Current billing period usage — how much of your $15 monthly credit you’ve used
- Top-up balance — remaining purchased credits
- Usage history — daily and weekly breakdowns
Tips for Efficient Usage
Choose the Right Model
Lighter models cost less per message. Use them for simple tasks and save premium models for complex reasoning. See AI Models.
Adjust Thinking Level
Lower thinking levels use fewer tokens. Use
/think off for quick lookups and /think high only when you need deep analysis.Compact When Needed
Long conversations accumulate context. Use
/compact to summarize and free up tokens. See Memory.Keep Memory Curated
A bloated MEMORY.md file loads every session. Keep it focused on facts your agent actually needs.