Every interaction processed by Novus AI uses resources from your configured AI provider and may also be subject to limits associated with your Novus AI plan.
Understanding these limits can help you configure your assistant efficiently and troubleshoot situations where AI responses are temporarily unavailable.
How an AI Request Works
When an eligible Discord message is received, Novus AI prepares the information required for the AI to generate an appropriate response.
Depending on your configuration, this may include:
The member's message
AI Instructions
Relevant Knowledge Base information
AI Routing instructions
Conversation context
Language requirements
Other information required for the response
Novus AI then sends the necessary request to your configured AI provider.
The provider processes the request using your selected AI model and returns the generated response.
Novus AI Limits vs. Provider Limits
It is important to understand that these are two separate systems.
Novus AI limits are determined by the Novus AI plan associated with your account or Discord server.
AI provider limits are determined by the external provider connected through your API credentials.
Reaching a limit on either service may affect your ability to generate AI responses.
Novus AI Plan Limits
Different Novus AI plans may provide different levels of access to features and usage.
Depending on your plan, certain functionality may be:
Included
Limited
Unavailable
Subject to usage allowances
Your current plan and its available features can be reviewed from your Novus AI account.
Plan features and limits may change as Novus AI services are updated.
AI Provider Usage
Your external AI provider may apply its own usage rules.
Depending on the provider and model, these may include:
Request limits
Token limits
Rate limits
Daily or monthly quotas
Model-specific limits
Account-level restrictions
Billing requirements
These limits are controlled by your AI provider and are separate from your Novus AI subscription.
API Costs
Some AI providers charge for API usage.
The amount charged may depend on factors such as:
Selected AI model
Amount of input processed
Amount of output generated
Number of requests
Provider pricing
Any charges from your external AI provider are separate from payments made for your Novus AI subscription.
Always review your provider's current pricing and usage dashboard for accurate billing information.
What Is Context?
Context is the information available to the AI when it processes a request.
A request may include more than the member's latest message.
For example, Novus AI may need relevant instructions, Knowledge Base information, routing information, or conversation history to understand the question and generate an appropriate response.
The amount of context an AI model can process is limited by the model itself.
Keep Your Configuration Efficient
More information does not always produce a better response.
Very large or repetitive AI Instructions can consume unnecessary context and make your configuration harder to maintain.
For better efficiency:
Keep general AI Instructions focused on behavior.
Store factual community information in the Knowledge Base.
Use AI Routing for topic-specific instructions.
Remove outdated or duplicate Knowledge Base content.
Avoid repeating the same rules across multiple configuration areas.
Use clear and concise instructions whenever possible.
A well-organized configuration helps Novus AI provide the AI model with more relevant information.
Avoid Processing Unnecessary Messages
If Novus AI is configured to process normal conversations in a busy Discord channel, your AI provider may receive requests that do not actually require AI assistance.
Consider using:
Dedicated support channels
Mention-only responses
Appropriate channel configuration
Clearly defined AI interaction areas
This can reduce unnecessary AI requests and provide a cleaner experience for your members.
Rate Limits
AI providers may limit how many requests can be processed within a certain period.
If many members interact with Novus AI at the same time, your provider may temporarily reject additional requests.
A rate limit does not necessarily mean your API configuration is incorrect.
The provider may simply require you to wait before additional requests can be processed.
Quota Limits
Some providers assign usage quotas to API accounts.
Once a quota is reached, AI requests may stop working until:
The quota resets
Additional usage becomes available
Billing is enabled or updated
The provider increases the account limit
Check your AI provider dashboard when you receive quota-related errors.
Model Limits
Different AI models may have different usage and context limits.
Switching models can change:
Available context
Response speed
API cost
Request limits
Response quality
Provider quota consumption
Always test your configuration after changing models.
If AI Responses Suddenly Stop
If Novus AI was previously working but stops generating AI responses, check:
Your Novus AI subscription and available plan features.
Your AI provider API key.
Your provider's remaining quota.
Provider billing status.
Rate limit errors.
Selected model availability.
Discord channel configuration and permissions.
Any recent changes to your AI configuration.
Identifying whether the issue comes from Novus AI, Discord, or your AI provider can help you resolve the problem more quickly.
Monitor Your Usage
If your AI provider provides a usage dashboard, review it regularly.
This can help you understand:
How frequently your assistant is being used
Which models are consuming API resources
Whether you are approaching provider limits
Whether your API usage is generating unexpected costs
For larger Discord communities, monitoring usage is especially important.
Build an Efficient AI Assistant
The goal is not simply to minimize AI usage.
The goal is to make sure each AI request has the information required to provide useful assistance without unnecessary processing.
A clear Knowledge Base, focused instructions, appropriate routing, and controlled response behavior can help you achieve this balance.
Next Step
Now that you understand the factors that can affect AI usage and responses, the final step in this section is verifying that your AI configuration works as intended.
Continue to Testing Your AI Configuration.