Usage
In this document, we'll walk you through the key metrics - tokens, conversations, and messages - that form the foundation of our service. You'll learn not only how these metrics are calculated but also practical strategies for managing them effectively to maximize the value of your subscription.
Understanding the Metrics
Tokens
Tokens are a part of your monthly allocation and are used every time a message is generated by the system, both for input and output. The number of tokens consumed depends on the model chosen - more complex models consume more tokens.
Tokens are calculated cumulatively within each conversation. For instance, if a user message consumes 10 tokens and the bot's response also consumes 10 tokens, the total is 20 tokens. If a subsequent user message consumes 10 tokens and the bot's response is another 10 tokens, the total is not 40, but 60 tokens (the sum of the first iteration of 20 tokens plus the second iteration that includes all messages thus far). You can control exactly how many messages are included in each iteration by modifying the model “Interaction Max Messages” settings.
Certain actions can lead to higher token consumption. For example, when using skillsets that include actions such as fetching a page, generating text, vision capabilities, etc. will consume more tokens. It's important to note that ChatBotKit is designed to optimize and consume as few tokens as possible.
Tokens are reset at the beginning of each billing cycle.
Conversations and Messages
Conversations and messages are directly tied to your monthly plan. The number of conversations and messages you create or send will increment these metrics on a 1-1 basis. Unlike tokens, conversations and messages are not cumulative.
Monitoring Your Usage
The usage metrics are updated in real-time with no delay. You can view these metrics at ChatBotKit Usage. Here, you can also download the complete raw log of all metrics, providing a detailed analysis of the usage metrics and reasons for their use.
Impact on Plan and Billing
When you approach your plan's limits, you'll receive notifications to help you stay informed. It's recommended to regularly check your usage statistics and plan accordingly to avoid any service interruptions.
Here are some key considerations for managing your plan usage:
- Monitor your daily and weekly usage patterns to identify peak usage periods
- Review historical usage data to make informed decisions about plan upgrades
If you consistently approach your plan limits, it may be more cost-effective to upgrade to a higher tier rather than risk service interruptions or overage charges.
Managing Your Usage Effectively
Choosing less complex (and thus cheaper) models can significantly reduce token consumption. Similarly, limiting the number of messages sent to the model through configuration options can also help manage your usage effectively.
By understanding these metrics and how they are calculated, you can make the most of your ChatBotKit subscription and ensure you're using our services in the most cost-effective way.
Conclusion
Understanding and effectively managing your ChatBotKit usage metrics is key to getting the most value out of your subscription. By monitoring your tokens, conversations, and messages, you can optimize your usage, control costs, and ensure a seamless experience. Remember to regularly review your usage dashboard, consider historical patterns, and make proactive adjustments to your plan as needed. With these best practices in mind, you'll be well-equipped to harness the full potential of ChatBotKit's powerful AI capabilities.