Amazon SP-API Rate Limits: The Complete 2026 Guide
Amazon's Selling Partner API powers every third-party tool. But between rate limits, throttling, and burst quotas, most developers spend more time fighting the API than building features. This guide covers everything you need to know.
Claude
Gemini
ChatGPTTL;DR - Key Takeaways
- •SP-API rate limits vary by endpoint, seller tier, and marketplace. The Orders API allows just 1 request/minute sustained.
- •Use the Reports API for bulk data. One report request can replace thousands of individual API calls.
- •DIY pipelines cost $300K-$500K to build and $100K+/year to maintain. Most teams underestimate by 6-12 months.
- •Nova's raw data service handles all rate limit complexity with hourly refresh to your warehouse.
Latest updates
Refreshed for 2026
- SP-API throttle behavior in 2026 still differs by endpoint. Orders, Reports and Finances each carry their own burst-quota limits and retry-after rules that any integration has to respect.
- Nova's managed ingestion runs zero-loss hourly pulls across 200+ Amazon metrics and 40+ fee types into the Seller Cockpit, from $29/mo or Custom.
- Related reading: Amazon Data API, Amazon data library and latest Amazon news.
Amazon's Selling Partner API (SP-API) powers every third-party analytics tool, repricing software, and inventory management system. But between rate limits, throttling, and burst quotas, most developers spend more time fighting the API than building features. This guide covers everything you need to know about SP-API rate limits in 2026.
If you've ever seen "Request too frequent" errors or watched your data pipeline grind to a halt, you understand the frustration. The SP-API isn't just one API. It's 20+ different endpoints, each with its own rate limit rules, authentication requirements, and quirks that Amazon rarely documents clearly.
We've spent years building data pipelines that handle hundreds of Amazon seller accounts. This guide shares everything we've learned about rate limits, throttling strategies, and why most DIY pipelines eventually fail. If you're evaluating whether to build your own pipeline or use a service like Nova's ready-made data delivery, this will help you make an informed decision.
What Are SP-API Rate Limits?
Rate limits control how many API requests you can make in a given time period. Amazon uses rate limits to protect their systems from overload and ensure fair access for all developers. When you exceed these limits, you get throttled (your requests fail until you slow down).
Requests Per Second (RPS)
The sustained rate at which you can make requests. Most endpoints allow 1-5 requests per second.
Burst Rate
The maximum requests you can make in a short burst. Allows temporary spikes above the sustained rate.
Restore Rate
How quickly your quota refills after being depleted. Typically measured in requests per second.
Daily Quotas
Some endpoints have additional daily limits, regardless of how slowly you make requests.
The Hidden Complexity
Rate limits vary by seller tier, marketplace, time of day, and endpoint. A limit that works fine for a small seller might throttle constantly for a large brand. Amazon also changes limits without warning, breaking pipelines overnight.
Rate Limits by Endpoint (2025)
Here's a comprehensive breakdown of rate limits for the most commonly used SP-API endpoints. These are based on Amazon's official documentation and our real-world experience.
| Endpoint | Rate (RPS) | Burst | Notes |
|---|---|---|---|
| Orders API | 0.0167 | 20 | 1 request per minute sustained |
| Reports API | 0.0222 | 10 | Report generation is async |
| Catalog Items | 5 | 40 | Relatively generous |
| Inventory API | 2 | 30 | FBA inventory updates |
| Finances API | 0.5 | 30 | Financial events & settlements |
| FBA Inbound | 2 | 30 | Shipment management |
| Notifications API | 1 | 5 | Subscription management |
| Product Pricing | 0.5 | 1 | Very restrictive |
| Advertising API | Varies | Varies | Separate from SP-API |
Pro Tip: The Reports API is Your Friend
Instead of making thousands of individual API calls to the Orders endpoint, request a report. One report can contain all your orders for a time period, drastically reducing API calls. The tradeoff is latency (reports take minutes to generate).
Skip the API Complexity
Nova delivers clean, analysis-ready Amazon data to your warehouse hourly. No rate limits to manage, no throttling to debug, no schema changes to chase.
Throttling Strategies & Retry Logic
When you hit rate limits, Amazon returns a 429 (Too Many Requests) error. Your code needs to handle this gracefully. Here are the strategies we use:
Request Queuing
Queue all API requests and process them at a controlled rate. This is essential when managing multiple seller accounts.
Multi-Account Complexity
If you're building for an agency or aggregator with 50+ accounts, each account has its own rate limits. You need 50 separate token buckets, 50 queues, and logic to prioritize which accounts get refreshed first. This is where DIY pipelines become unmanageable.
"We spent 8 months building our own SP-API pipeline. By month 4, we had working data. By month 8, we realized maintenance was eating 40% of one engineer's time. Switching to Nova was the best decision we made."
Why DIY Amazon Data Pipelines Fail
We've talked to dozens of teams who tried to build their own SP-API pipelines. Here's what typically goes wrong:
Underestimating Scope
Teams estimate 2-3 months. The reality is 12-18 months for a complete pipeline. Rate limit handling alone takes weeks to get right.
Schema Changes Break Everything
Amazon changes field names, adds new fee types, and deprecates endpoints quarterly. Each change requires engineering time to fix.
Data Reconciliation Nightmares
The Orders API doesn't match the Reports API which doesn't match Settlement Reports. Figuring out which source is "correct" takes months.
Ongoing Maintenance Cost
Even after launch, plan for 0.5-1 full-time engineer just to keep the pipeline running. That's $100K+/year in perpetuity.
DIY Initial Cost
$300K-$500K
6-18 months development
DIY Annual Maintenance
$100K+
0.5-1 FTE dedicated
Nova Time to Data
24-48 hrs
Full historical backfill
Struggling with Amazon Data?
We've solved these problems for 500+ brands. Stop wrestling with APIs and get clean, query-ready data delivered to your stack.
How Nova Handles Rate Limits
Our ready-made Amazon raw data service Abstracts away all the complexity of SP-API rate limits. Here's what we do behind the scenes:
Intelligent Scheduling
We optimize request timing across all accounts to maximize throughput while staying under limits.
Automatic Retries
When throttled, we automatically back off and retry. You never see the failures.
Report-First Strategy
We use bulk reports instead of individual API calls wherever possible, reducing total requests by 99%.
Hourly Refresh
Despite all the rate limit complexity, we deliver fresh data to your warehouse every hour.
What You Get with Nova's Raw Data Service
Clean, normalized data: 200+ report types merged into unified schemas
Pre-calculated KPIs: 200+ metrics ready to query
Historical backfills: 2+ years of historical data included
Multi-marketplace: US, UK, DE, FR, and more in one dataset
BigQuery or Snowflake: Data delivered to your warehouse
Frequently Asked Questions
Next Steps
If you're evaluating how to get Amazon data into your analytics stack, you have three options:
Build It Yourself
If you have 12+ months of engineering time, dedicated data engineers, and budget for ongoing maintenance. Good for unique requirements that no service can meet.
Use Nova's Raw Data Service (Recommended)
Get data in your warehouse in 24-48 hours. We handle rate limits, schema changes, and maintenance. Best for teams that want to focus on analysis, not infrastructure. Learn more →
Use Seller Central Reports
Download CSV exports manually. Works for small sellers with simple needs. Doesn't scale and data is always 24-48 hours stale.
Related reading: How to Get Amazon Seller Data into BigQuery | Snowflake Guide | Openbridge Alternatives
Skip the Pipeline Build
Get normalized Amazon data delivered to your warehouse in days, not months. 200+ pre-calculated KPIs, hourly refresh, zero maintenance.
Continue Learning
Explore more expert insights to grow your Amazon business
Amazon Seller Data to Snowflake
Your data team already uses Snowflake. Getting Amazon seller data in shouldn't require 18 months of engineering. Learn 3 paths: DIY SP-API, ETL tools, or Nova's pre-built Snowflake delivery with hourly refresh.
Amazon Seller Data to BigQuery
Building an Amazon data pipeline from scratch takes 18+ months and costs $2M+ in engineering time. Learn three paths to getting Amazon seller data into BigQuery: DIY with SP-API, ETL tools like Airbyte, or pre-built solutions. Includes SQL query examples for P&L analysis.
Best Openbridge Alternative 2026 - Amazon Data for BigQuery
Openbridge pioneered Amazon SP-API data pipelines for analytics teams. In 2026, alternatives ship ready-to-query per-SKU profit instead of raw reports. This guide compares Openbridge to Nova's ready-made Amazon data service.