Tiered Usage & Cost Limits
Set granular usage quotas and spending caps based on user attributes like subscription tiers to enforce budgets and fair usage.
Define maximum monthly estimated LLM spending limits for individual users or entire tiers to prevent budget overruns.
Allocate different request allowances (RPM, monthly, etc.) to distinct user groups (e.g., Free vs. Pro vs. Enterprise).
Automatically apply the correct limits by matching user attributes (e.g., subscription plan, role from JWT/profile).
Enforce fair usage policies, manage resource allocation effectively, and align costs directly with user value.
Enforce Budgets and Fair Usage with Tiered Limits
Not all users are created equal. Prompt Shield allows you to move beyond one-size-fits-all protection by defining distinct usage limits and cost controls based on user attributes, such as their subscription plan, role, or organization ID. This ensures fair resource allocation, enables effective monetization strategies, prevents abuse from lower-value tiers, and keeps your LLM costs predictable across your entire user base.
Key Capabilities
Defining Tiers and Identification
- Flexible Attribute Mapping: Configure Prompt Shield to identify a user’s tier based on specific attributes found in their JWT claims, user profile data, or API key metadata (e.g.,
user.subscriptionTier,org.plan,apiKey.tier). - Multiple Tiers: Define as many tiers as needed (e.g., “Free”, “Basic”, “Pro”, “Enterprise”, “Internal”).
- Default/Fallback: Set default limits for users who don’t match any defined tier attribute.
Setting Rate Limits per Tier
- Multi-Window Limits: For each defined tier, configure specific request limits across various time windows (Requests Per Minute, Hour, Day, Month).
- Concurrency Control: Limit the number of simultaneous requests allowed per user within each tier.
- Granular Control: Ensure free users have appropriate limits while providing higher throughput for paying customers.
Setting Cost Caps per Tier
- Monthly Budget Enforcement: Set a maximum estimated LLM cost allowed per user (within a specific tier) per month.
- Prevent Individual Overspending: Cap the potential cost impact of any single user, regardless of their request volume, ensuring budget predictability.
- Align Cost with Value: Ensure that higher-value subscription tiers can access more resources while keeping costs for lower tiers strictly controlled.
How It Works
- Request Received: Prompt Shield intercepts the request before it hits your core application or the LLM API.
- User Identification: Extracts the relevant user identifier (User ID, API Key, etc.).
- Tier Lookup: Retrieves the configured tier attribute associated with the identified user (e.g., fetches
user.subscriptionTier). - Limit Application: Applies the specific rate limits and checks against the cost cap defined for the user’s matched tier (or the default limits if no tier matches).
- Enforcement: Allows or blocks the request based on the applicable limits for that specific user and tier.
Benefits
- Predictable Costs: Cap spending per user/tier, eliminating surprises.
- Fair Usage: Allocate resources appropriately based on user value or plan.
- Monetization Enablement: Directly tie API usage allowances to subscription plans.
- Reduced Abuse: Prevent low-value or free tiers from consuming excessive resources.
- Scalable Control: Manage limits effectively across a diverse user base.
Implement Tiered Usage & Cost Limits with Prompt Shield to bring predictable, value-aligned control to your LLM application’s usage and spending.
See it in action
See how easily you can define and manage different limits for your user tiers in Prompt Shield.
