In general, Rate limits are restrictions that our API imposes on the number of times a user or client can access our services within a specified period of time. Datura API enforces rate limits to ensure fair usage and maintain system stability. The limits vary based on the user’s account tier and are subject to change based on agreements.

​

Why Datura’s API Rate Limits?

Rate limiting is a fundamental control mechanism in API architectures designed to ensure service stability, optimize resource utilization, and protect against misuse. Its implementation serves several critical purposes:

​

Mitigation of Malicious Activity:

Rate limiting acts as a defensive barrier against various forms of abuse, such as distributed denial-of-service (DDoS) attacks and other forms of automated request flooding. By capping the number of requests a client can issue within a specified time interval, the API prevents a single actor from monopolizing the system’s processing capacity, thereby safeguarding against resource exhaustion and service interruptions.

​

Fair Resource Allocation:

In multi-tenant and high-traffic environments, rate limiting enforces equitable access to API resources. By throttling excessive usage on a per-client basis, it prevents any one client from consuming disproportionate amounts of bandwidth or compute resources. This ensures that all authorized users receive consistent and predictable service levels, contributing to overall system fairness and compliance with service-level agreements (SLAs).

​

Infrastructure Load Management:

Rate limits help manage the aggregate load on backend systems. By regulating the rate of incoming requests, the API infrastructure can maintain optimal performance, prevent congestion, and avoid processing bottlenecks during peak traffic periods. This controlled throttling of traffic not only improves response times but also enhances the resiliency of the service under variable load conditions.

​

Rate Limiting Metrics and Enforcement

API rate limits within Datura’s API are enforced using multiple dimensions to control request volume and resource consumption effectively. These limits include:

• RPM (Requests Per Minute): The maximum number of API calls a client can make within a one-minute window.
• RPD (Requests Per Day): The total number of API requests allowed within a 24-hour period.
• RPMn (Requests Per Month): The maximum number of API calls a client can make within a 30 days.

These thresholds ensure optimal API performance and fair resource distribution. Rate enforcement follows a first-limit-reached approach, meaning API access is restricted once any of the defined limits is exceeded. For instance, if an API client submits 10 requests per minute, it will be temporarily throttled even if the daily or monthly limits remain unmet. This mechanism helps prevent service disruptions, mitigates abuse, and maintains system stability by managing overall infrastructure load efficiently.

​

Desearch API Rate Limits

Each tier of the Desearch API has specific rate limits for RPM, RPD, and RPMn. These limits are designed to accommodate a range of usage scenarios, from light to heavy traffic, while ensuring optimal performance and reliability. The following table outlines the rate limits for each tier:

​

Handling Rate Limits

To optimize API usage and prevent hitting rate limits:

✅ Monitor Usage: Regularly track your API usage via the Datura Console.

✅ Implement Exponential Backoff: If you encounter rate limit errors, use a retry mechanism with incremental wait times.

✅ Upgrade Your Plan: If you require higher limits, contact support@datura.ai for plan upgrades.

✅ Optimize API Calls: Minimize redundant requests and batch queries when possible to stay within limits.