Skip to main content

Introduction

To ensure the stability and reliability of our platform, we have implemented rate limits across all API endpoints. These limits are designed to prevent abuse and ensure fair usage for all users. Below, you will find a detailed breakdown of the rate limits applied to each service. If you require higher rate limits or have specific use cases that exceed these thresholds, please contact our support team to discuss potential solutions or upgrades to your plan.

How Rate Limits Work

Each plan has two limits, and a request must pass both:
  • Burst limit: How many requests you can send in quick succession. Think of it as a bucket that holds a fixed number of requests and drains at 1 request per second. If the bucket is full, the request is rejected. A moment later there is room again.
  • Daily limit: How many requests you can send in a 24-hour window. The window starts with your first request and resets 24 hours later.
If you exceed either limit, you will receive a 429 Too Many Requests response with one of these messages:

Limits by Plan

What Each Limit Applies To

Limits are counted separately for each service, so heavy traffic on one doesn’t use up another. For example, on the Pro plan each of your public silos can serve 10,000,000 requests a day on its own.

Rate Limit Headers

Successful responses include headers that show where you stand against the daily limit.
The prefix depends on the service:

Free Trial Limits

The free trial has one extra limit: 1,000 requests in total, across your silo URLs, the Manage API, and MCP. Unlike the daily limit, this count never resets. While you are on the trial, responses include these headers:

When the trial ends

Once the 2 days are up or the 1,000 requests are used, requests return 402 Payment Required with this body:
reason is time when the 2 days have passed, or cap when the requests ran out. Choose a plan to bring your silos back online.