🔢 Understanding Tokens

Every time you interact with Blockli Assistant, the system processes your request using a powerful AI model. To ensure fairness and cost-efficiency, we measure usage based on tokens—the building blocks of every AI conversation.

A token is a small unit of text. Tokens can be as short as one character or as long as one word. For example:

• The word hello is one token

• The phrase How are you today? is about 5 tokens

Tokens are how AI models read and process language, and both the input (your prompt) and the output (the assistant’s response) count toward your usage.

Blockli Assistant usage is measured using two token types:

1. Prompt Tokens

   These are tokens used when you send a message to the assistant. They include your message plus any system context or data the assistant needs to understand your question.

2. Completion Tokens

   These are tokens used when the assistant generates a response to your message.

🧾 Total token usage per interaction = Prompt Tokens + Completion Tokens

Your token usage is tied directly to your credit balance. Here’s how it works:

• Each time you send a message and receive a response, the total number of tokens is calculated.

• The cost of those tokens is subtracted from your available credits.

• When your credits run low, you’ll be prompted to top up or add more.

🔄 Want to avoid interruptions? Keep an eye on your usage in the Usage Metrics section of your customer portal.

Several factors can influence how many tokens are used:

• Length of your message (longer questions use more prompt tokens)

• Size of the assistant’s reply

• Contextual data (e.g., analytics, user activity summaries) included in the prompt

• How specific your question is (more precise = fewer tokens)

• Keep your questions concise

• Use assistant summaries instead of raw data dumps

• Leverage built-in analytics instead of asking for full reports

Our system is already optimized to use precomputed insights and efficient prompt engineering to help reduce token usage behind the scenes. You get the insights—without unnecessary costs.

Head to your Usage Metrics page in the Customer Portal to see:

• Total tokens used

• Prompt vs. completion breakdown

• Credits consumed

• Remaining balance